a eres ang Pibrary of the Museum OF | COMPARATIVE ZOOLOGY, AT HARVARD COLLEGE, CAMBRIDGE, MASS. The gift SEE eer eR viet Mo, 44 44 Stes Mf, 184 lo: Pl ($93. — TER AS CIE INTs OF THE AMERICAN PHILOSOPHICAL SOCIETY, ISL IDI ID ANG IP ls Obee Oo IIe elias FOR PROMOTING USEFUL KNOWLEDGE. VOL. XVILU.-NEW SERIES. IPWIBILUSIEUIID) IBY WMSa) SOCIMUIA We ‘Philadelphia: MACCALLA & COMPANY INC., PRINTERS, "1896. COWEN aS (Or ViOliee Var d= JA 1 On ae = ARTICLE I. Old Babylonian Inscriptions Chiefly from Nippur. By H. V. Hilprecht, Ph.D. : 6 5 ; : 4 5 d= JX dsh'n Abae = ARTICLE II. The Mammalia of the Deep River Beds. By W. B. Scott . 9 ¢ : : ‘ 0 . F 3 : 55 ARTICLE III. The Classification of the Ophidia. By E. D. Cope 3 é 4 5 : 0 : 5 : ¢ 9 : 186 I>) A, Igy Ge) ALICE ARTICLE IY. Old Babylonian Inscriptions Chiefly from Nippur. Part II. By H. V. Hilprecht, Ph.D. ; ; ; : 221 , Bante a3 5 We BON TS AOS OF THE AMERICAN PHILOSOPHICAL SOCIETY. CORRECTIONS. Page 9,1.29: For Allen read Allan. Page 14,1. 31: For lissukd read lissuhd. Hae 1, I Be oan 5 is to be removed. Page 29,126: ForIread II. slowing Page 87,1. 5: For Barnaburiash read Burnaburiash. Sa ouved fa Page 43, 1. 26: For Ménaut read Ménant. oration liversity +, as the 3 labors. tly date, ngs and results. In the meantime for the student 1 nave appenucu—vw ee a1... Uction a Bibliography of those contributions of its members to various periodicals which relate to its work. Towards the close of the year 1891 there arrived at the Museum of the Univer- sity some eight thousand clay tablets, together with several hundred fragments of vases and other inscribed objects in stone, which had been disinterred in Nippur or Nuffar.* Iwas able at once to proceed with the work of cleaning and examining ® This is the present designation of the extensive ruins by tne Affek tribes, in whose territory they are situated. Although I repeatedly had the Arabs of the neighborhood pronounce for me the name they give to the ancient Nippur, I never heard from their lips the pronunciation Niffer, to which Layard and Loftus have given currency among Assyriologists. A. P. S:—VOL. XVIII. A. TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY. oo ARTICLE I. OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. BY H. V. HILPRECHT, Pu.D., Professor of Assyrian and Curator of the Babylonian Museum in the University of Pennsylvania. Read before the American Philosophical Society, November 4, 1892. PREFACE. Tue old Babylonian Cuneiform Texts, which are published in the following pages, are a part of the harvest gathered by the Expedition sent out in the summer of 1888, under the auspices of the University of Pennsylvania, for the exploration of Babylonia. The Rey. Dr. John P. Peters, Professor of Hebrew in the University of Pennsylvania, was the Director of the Expedition, while the subscriber, as the Assyriologist of the University, accompanied it during the first year of its labors. As the history of the Expedition is to be published by its Director at an early date, I here abstain from giving any account of its origin, members, undertakings and results. In the meantime for the student I have appended to the Introduction a Bibliography of those contributions of its members to various periodicals which relate to its work. Towards the close of the year 1891 there arrived at the Museum of the Univer- sity some eight thousand clay tablets, together with several hundred fragments of vases and other inscribed objects in stone, which had been disinterred in Nippur or Nuffar.* I was able at once to proceed with the work of cleaning and examining * This is the present designation of the extensive ruins by the Affek tribes, in whose territory they are situated. Although I repeatedly had the Arabs of the neighborhood pronounce for me the name they give to the ancient Nippur, I never heard from their lips the pronunciation Niffer, to which Layard and Loftus have given currency among Assyriologists. IXo 124 SE AVVOMW HAVING ZN 6 OLD BABYLONIAN INSCRIPTIONS them. Three months later I had obtained a general idea of their contents and their age, and had catalogued about a third of them. On the basis of a report submitted to the Publication Committee of the Expedition, of which Mr. Clarence H. Clark is Chairman, a plan was carefully devised for making these cuneiform inscriptions accessible to a wider circle of students, with as much speed and method as possible. With this view the Assyriologists of America and Canada were invited to lend their aid to the preparation of an extensive work on the Expedition and its results. A number of them haye given assurance of their readiness to do so. In April, 1892, the undersigned was entrusted by the Committee with the edit- ing of the series containing the Cuneiform Texts, and, at the same time, was requested to undertake at once the preparation of the first volume of these texts. It is esti- mated that the series will extend to eight or possibly ten volumes. Their general plan and character are well explained in a report submitted to the American Philo- sophical Society by a special committee, of which Mr. Talcott Williams was the Chairman, at the stated meeting of May 20, 1892. I take this opportunity to acknowledge the liberality of the venerable American Philosophical Society of Philadelphia, as shown in the promptness with which it has undertaken the publication of the present volume, by giving it a place in its learned and valuable Transactions. I hope that in the future the Society will continue to evince its interest in making such labors accessible to the republic of letters, by ex- tending its sympathy and support to the undertaking whose plan has been described. A word more must be said as to the manner in which it is intended to prepare the Cuneiform Texts for the use of the Assyriologist. For the sake of securing uniformity throughout the series, and of avoiding what would make it excessively costly, it was necessary to reproduce the inscriptions by photograph from copies made by hand, rather than from the objects themselves. Besides, the editor some time ago reached the conclusion that the method of direct photography is not at all satis- factory in the case of many inscriptions. The best which has been done by that expensive process is beyond question the work edited by Ernest de Sarzec and Léon Heuzey under the auspices of the government of France: Découvertes en Ohaldée. Tt possesses unique merits. But in spite of all the care that has been taken to secure an exact reproduction of the monuments, any Assyriologist who has worked through such texts as are found on Plates 33, 35 and 41, No. 1, will agree with me that the decipherment, especially of the margins, makes a very severe demand upon the eye- sight—a circumstance which makes the prompt and comprehensive use of the con- tents of this beautiful work sometimes difficult. After mature consideration, there- fore, the Committee found it most suitable to reproduce the Cuneiform Texts from CHIEFLY FROM NIPPUR. 7 copies made by the hand, and to employ photographs from the objects themselves only occasionally, to enable the Assyriologist to verify the copies and to perceive the archeological character of the inscribed objects. The first volume, whose first part I publish herewith, contains only inscriptions in old Babylonian which have been found on vases, door sockets, stone tablets, votive axes, bricks, stamps, clay cylinders, and similar objects of a monumental character. As the most of them belong to that period of Babylonian history of which our knowledge is very defective, the most painstaking care has been applied to auto- graphically reproducing the originals with the utmost faithfulness. The editor has kept in view, not only the making fresh and important materials accessible to students of Assyriology, but also the doing his part in placing Babylonian paleography on a better foundation. For this end every text has been reproduced in its actual size and form—that is, so as to show all the peculiarities of the scribes, not only as to the dimensions, shape and position of every character and group of such, but also their distance from one another, as was so admirably done by Sir Henry Rawlinson and Edwin Norris in the first volume of The Cuneiform Inscriptions of Western Asia. The investigations and collections I have made since the year 1883, and my lec- tures regularly held since 1886 on “The Development of Cuneiform Writing in Babylonia and Assyria,” have led me to conclude that the size and relative position of individual cuneiform characters, and certain combinations in which they frequently occur, have been a factor of importance in the development of the stereotyped forms of later date. The detailed proof of this I must reserve for the present until more urgent matters have been disposed of. At any rate, careful editions of texts, and a faithful reproduction of the peculiarities of the individual Babylonian scribe, have become a pressing necessity for the progress of Assyriology, if we are to attain in this field anything like the results which Huting has achieved in other departments of Semitic paleography, and which are so necessary in determining the age of frag- mentary and undated inscriptions. In spite of the scantiness of representative old Babylonian texts of which the Assyriologists could make use, it would not have been possible for them to have differed by 500, 1000 or even 2000 years as to the date of inscriptions, if such texts had always been reproduced carefully for their use. It is to be expected that the excavations still proceeding at Nuffar will supply the completion of texts here given in fragmentary shape, and that several finds will make their way into various Huropean and American museums by reason of the ’ thievishness of the Arabs employed in them, who also may carry on excavations on their own account.* For this reason I have shown as exactly as possible the fracture * Cf. my note in Zeitschrift fur Assyriologie IV, p. 282 seg. Sayce, Records of the Past?, Vol. III, pp. x, note SUNG 8 OLD BABYLONIAN INSCRIPTIONS of such fragments. Tt was thus that I myself, after the printing had begun, was enabled to recognize the connection of Pl. 21, No. 41 and No. 46, and between Pl. 22, No. 50, and Pl. 26, No. 74. Where I have shaded the inscription in my copy, it is not meant to indicate that the reading is to me uncertain, but that it can be recognized only in a special light and by a practiced eye, looking at it from an especial angle. How necessary it was to make an autograph copy of such inscriptions may be seen by comparing PI. 23, Nos. 56, 57, and the direct photographic reproduction on PI. X. A restoration of broken ’ characters and lines I have avoided on principle, even when there was no doubt in my own mind as to what was missing. My translations will show in due time what my understanding of such passages is. For obvious reasons, I have given the characters in some inscriptions only in outline. Of the plates which reproduce the inscription on the Abu Habba slab I have avoided altogether making an autograph copy, since I thought this needless. This stone was found in Abu Habba during the excavation undertaken at the private expense of the Sultan in 1889, and is now in the Imperial Museum at Constantinople. Through the courtesy of His Excellency Hamdy-Bey, a cast of it was furnished to our Expedition. Unfortunately this was broken in pieces in transportation, but it was restored by one of my students. It is this cast that has been directly photographed for the present publication. Some portions of its margin have an indistinctness, which is faithfully shown by the photographic re- production. To convey to scholars a clearer picture of the ruins of Nippur, and to show the sites at which the several inscriptions were found, a plan of the excavations of the first year is given. In the Table of Contents the texts are described with reference to this Plan, which has been prepared in accordance with the bas-relief of the ruins made by Mr. Charles Muret in Paris under the supervision of Mr. Perez Hastings Field, the architect of the Expedition. In determining the mineralogic character of the several stones, I have had-the assistance of my colleagues, Drs. G. A. Koenig and H. Smith, of the University of Pennsylvania, to whom I extend my thanks. As I was able to accompany the Expedition only during the first year, I am greatly indebted to my esteemed col- league, Dr. Peters, for much valuable information as to the sites in which objects were found, and for sketches and copies of a series of objects and inscriptions which he made during its second year. As the antiquities disinterred arrived in this coun- try at long intervals, I found myself obliged to proceed with the help of casts, squeezes, electrotypes and Prof. Peters’ notebooks, in order not to delay needlessly the publication of the Texts. This circumstance, however, prevented my determining CHIEFLY FROM NIPPUR. 9 at the outset the material of the whole volume. At the opening of each new box I found myself compelled to withdraw some pages and substitute others, until the commencement of the printing, in October of last year, made further alterations and a more systematic arrangement impossible. The second part of this volume, which will appear in about half a year, will furnish further inscriptions of kings who are already represented in the first. Nor will it be possible entirely to avoid this defect of arrangement in other volumes, so long as the excavations at Nippur continue to bring to light new inscriptions of the same rulers. If, however, we were to delay the publication of the inscriptions until the complete results of the systematic explorations of the ruin-heaps at Nippur were at hand, it would have been necessary, according to my careful calculation, to wait some twenty years, sup- posing that the excavations were pushed forward with a force of some hundred Arab workmen. On account of its importance and its close connection with the class of Cas- site votive inscriptions here published, I have included the cuneiform text on the lapis lazuli dise of King Kadashman-Turgu, which probably came from Nippur,* and is now in the Museum of Harvard University,t Cambridge, Mass. Prof. D. G. Lyon kindly gave me leave to publish this, and placed at my disposal a cast of the disc, for which he has my warmest thanks. The transcription of the names of kings in the Table of Contents is the usual _ one. Strassmaier, Mabon. 133, 4. 6 Strassmaier, Vabon. 132, 4. Of. Peiser, Aus dem Babylonischen Rechtsleben I, p. 11. ‘The same principle of abbreviating names in everyday use occurs among nearly all ancient nations. Cf. e. g., Erman, Agypten und Ayyptisches Leben im Altertum, p. 233 ; also the Hebrew dictionaries ; Fick, Die griechischen Per- sonnenamen ; O. Crusius, Newe Jahrditcher, 1891, pp. 385-894: ‘Die Anwendung von Vollnamen und Kurznamen bei derselben Person.’’ For the last two references I am indebted to my friend and colleague, Prof. W. A. Lamberton. 5 Shargant, ‘the powerful.” See p. 18, note 4. ® Hommel, Gesch., p. 301. 0 P. 8. B. A., VII, p. 67 seq. 1 Cf. Hommel, J. ¢., p. 303. A. P. §.—VOL. XVIII. C. 18 OLD BABYLONIAN INSCRIPTIONS 3. It is absolutely impossible to regard Sargon, father of Naram-Sin, as “perhaps an invention of legend.”' But were he one of the best known and mightiest rulers of the olden time,” it was to be expected that some monuments of his would be found in the thorough exploration of the ruins of the temple at Nippur, where the greatest number of texts of his time* ever found has been brought to light. Where inscriptions of his less known son Naram-Sin, and of the hitherto altogether unknown Alusharshid, have been discovered, it was @ prior’ probable that inscriptions of Shargina = Shargéna = Shargani(a) would also come to light. Therefore the very absence of the name in the inscriptions there discovered is, in itself, a proof that the ancient king whose name commences with Shargdnz, and who is represented by six inscriptions, is no other than Sargon, the father of Naram-Sin. From this it follows naturally that the later Shargzna was merely an abbreviation of Shargani-shar-dla. According to Oppert, the name signifies ‘‘ mighty is the king of the city.” * There were also found in Nippur two brick stamps of Naram-Sin, son of Sargon I. Both contain the same legend. The moulds, however, that were used in making them differ slightly in size and shape. The inscription reads: 1. ““Nardm-""Sin 2. bint 3. bit “Bel, “Naram-Sin, builder of the temple of Bél.” If we may base an argument on the place in which the stamps were found, as to the location of Naram-Sin’s building, we might conclude that he built a shrine immediately on the canal south from the Zgqurratu, whilst his father confined himself in his building to the east side of the temple platform. In any case, from the contents of the 1 Winckler, Gesch., p. 39. » As is proved by the inscriptions of Nabfina’id, where he is called ‘‘king of Babylon’’, by the ‘‘ Legend of Sar- gon,’’ the Tablet of Omens IV R. 34, and the mention of his name in the List V R. 44, 18, a, }. Hommel, who reads erroneously Lugal-girinna (1. ¢., pp. 801, 307, note 4) in the last quoted passage, distinguishes Sargon of the list as Sargon IT, c. 2000 B.C., from the ancient Sargon I. His arguments are not convincing (cf. also Winckler, Unters., p. 45, note 2). It is especially ‘the historical background of the work ’’—the mention of Elam, Guti, etc., at such an early period, which is the most valuable evidence for the high antiquity and reliability of the statements contained in the astrological work. Cf. my remarks in connection with the inscriptions of the king of Guti and Alusharshid. Six inscriptions of Shargdani-shar-ali, two of Nardm-Sin, and sixty-one inscribed vases (or fragments) of Alusharshid, “Z. A. III, p. 124. Of. V RB. 41, 29a. b.: shar-ga-nu=dannu. Shargdnw is u noun formation in an (Delitzsch, Gram., 4 65, No. 35) from a root shardgu, which seems to mean ‘‘ to be powerful, mighty.’’ Cf. the Hebr. proper name iW. Likewise the names Bingani-shar-ali and Al-usharshid contain the formative element lw. There are reasons for identifying this lw (Alw) with Alw ki, used as an ideogram for “ Babylon’’ by Nebuchadrezzar IT (misunderstood by Delitzsch, Worterbuch, p. 6). Cf. Hilprecht, The Sunday School Times, 1892, No. 20, p. 306 seg. Nebuchadrezzar uses even mahazu alone (urbs) for ‘‘Babylon.” Cf. ¢. g. VR. 34 (Z. A. IL, p. 142-44), col. I, 13: zanan mahazi, “to adorn the City’”’ (2. e. Babylon, not ‘die Stadte,’’ Winckler in Schrader’s A. B. III, Part 2, p. 39). For the use of Alu without i, cf. below Kish (Kishshatu). CHIEFLY FROM NIPPUR. 19 inscriptions of Sargon and Naram-Sin it follows that the dominions of both included Nippur.’ The list of ninety-two garments, Pl. 6, was found near the inscriptions of Naram- Sin. As it is written in Semitic (ef. 1. 6, rabdtum), and as, paleographically, there is no objection to such a conclusion, it belongs probably to Naram-Sin, or, in any case, to one of the earliest Semitic kings of Babylonia. In this connection, I call attention to the interesting and important fact that the fragment of another vase (or probably of several) was discovered in the same deep-lying stratum as the inscriptions of Sargon and Alusharshid, and close by them. This fragment” contains the statement that ‘“ Hn-te(men)-na, patesi® of Shirpurla,” presented the vase to Bél of Nippur. When to this we add that a vase of Naraim- Sin,‘ and another of Alusharshid, as I have been informed, was found in Tello, we may safely conclude: 1. That the dominion of Sargon,’ Nardm-Sin and of their immediate successors (or predecessors’ ) extended also over the whole of South Babylonia ° (at any rate, as far as Shirpurla ‘). 2. That the chronology of the oldest Semitic rulers of Babylonia is approximately the same* as that of the earliest patesis of Shirpurla. 3. That the “kings of Shirpurla” are earlier than Sargon (or Alusharshid °). It was apparently Sargon I or Alusharshid who put an end to the independence of the kingdom of Shirpurla. This is not the place for a detailed statement of all my reasons. ‘They will be found in full elsewhere. To the early Semitic rulers of Babylonia already known must now be added, in consequence of the discoveries at Nippur, King URU-MU-USH, as his name is written. Not less than sixty-one fragments of different vases of his have been excavated from the temple. : As to the material of the vases cf. Table of Contents. The fact that they were found close to the monuments of Sargon, that like them they are written in Semitic, that the phraseology of Pl. 4, 1. 11, 12 is very similar to lines 6, 7 of the vase inscrip- 1 Cf. above, p. 15, note 5, and p. 25, note 3. 2 It will be published in Vol. I, Part 2. 3 T hold that the change of the title of Jugal into pates? in the case of the princes of Shirpurla is an indication of their political dependence (Hommel, J. ¢., p. 296). Jensen’s view (Schrader’s A. B. III, Part 1, pp. 6-8) is some- what different. 4 According to Oppert. Cf. Hommel, Gesch., pp. 299, note 1, 309. 5 See my remarks in connection with the texts of Alusharshid. ®° Cf. Hommel, J. c., pp. 296, 311. : 7 Winckler’s suggestion that Shirpurla is not identical with the modern Tello or part of these ruins ( Gesch., pp. 24, 31, note 1, 44, 326), but that it lay in North Babylonia, is quite improbable, to me even impossible. 8 In this I slightly differ from Hommel (J. c., p. 296), who places Sargon and Naraém-Sin a little later than the oldest patesis of Shirpurla. 20 OLD BABYLONIAN INSCRIPTIONS tion of Naram-Sin, that paleographically they show the characteristic features of the inscriptions of Sargon and his son, all this points to the first half of the fourth mil- lennium as the approximate date when they were written. As the language of the inscriptions is Semitic, I regard the name of the king also as Semitic and read ten- tatively Alu-usharshid,! 7. ¢., “He (some deity) founded the city.” ” The discovered inscriptions of this king may be classed in four groups, consist- ing of thirteen, eleven, six and three lines respectively. Only three of the three line legends’ have been preserved intact. Though not a single complete text of the six- line inscriptions has been excavated, yet the faint traces to be seen in the third-line of Pl. 1V, No. 18, and the space left for the restoration of the text, justify my read- ing of Pl. 5, No. 6,1. 1-3. The fragment reproduced on PI. 5, No. 10, is the only remnant of an eleven-line inscription found at Nippur. It is in all respects simi- lar to the thirteen-line inscriptions, with this difference only that 1. 11, 12 of the latter, on namrak Hlamti ", were omitted. The inscription of thirteen lines has been reconstructed from eleven fragments, three of which (PI. II], Fragm. 8891, 8892, a, b) belonged to a large dolomite vase and formed the basis of my text. Highteen fragments of all the excavated vases may confidently * be referred to this group. The long inscription, of which some of the shorter ones are possibly abbreviations,’ reads : 1. A-na 2. Bél 8. Alu-usharshid 4. shar 5. Kishshatu 6. i-nu 7. Hlamtu™ 8.2% 9. Ba-ra~-se™ 10. inira 11. in nam-ra-ak® 12. Hlamti™ 13. iddin (A-MU- 1Cf. Briinnow, J. c., 5082, 5068. ? Cf. Hilprecht, Z. A. VII, p. 315, note 1, and Pinches, The Academy, September 5, 1891, p. 199. Even if the name be transliterated Urumush, it may be Semitic. In this case the Orchamus of Ovid (Metam., 4, 212) offers itself for com- parison. 3 In spite of their identical contents I reproduced two of them (Pl. 5, Nos. 7 and 8), because of the slight differ- ence in the form of the characters USH and sharru, and because we do not possess a superabundant supply of texts dating from that ancient period to which they belong. The sign published on Pl. 5, No. 9, and resembling the Old Babylonian character for dlu, ‘‘god,’’ is found on the bottom of a third vase of the three-line group, and is, no doubt, merely a ‘‘trade-mark.”’ *T include here only those fragments of which portions of 1. 5-13 have been preserved. Some of the other frag- ments, however, probably belong to the same group. 5 Necessary because of limited space. 5 This word has been variously translated. Tiele (@esch., p. 115) and others before and since changed namrak into Apirak, a city mentioned on the tablet of omens, col. II, 12-14. Hommel (Gesch., pp. 279, 309) translates it ‘‘ polished work,”’ whilst Winckler (Gresch., p. 88) is content to render it simply ‘‘work.’’ But all this is mere guess work. To my knowledge, the word has been found thus far only in three passages, in the above text of Alusharshid, on the vase of Naram-Sin and in Gudea B, col. 6, 66. In the last passage we read 1. 64-69: gish KU wwAn-sha-an Nima ki mu-sig nam-ra-aga-bi Ung Nin-gir-su-ra H-ninnh-a mu-na-ni-tur, ‘With (his) weapon he smote the city of Anshan in Elam, brought its spoil into Eninnfi to Ningirsu.”’ Cf. Jensen (K. B. III, Part 1, pp. 88, 39) on this passage. The latter’s hesitation about the reading Vima ™, “Elam” (exactly so written above), and the meaning of namrak is unnecessary. As early as eight years ago, Amiaud, with his wonted insight, conceived the correct meaning of the word (Z. K. I, p. 249). Whether it is Sumerian or Semitic remains to be determined. As we do not possess long CHIEFLY FROM NIPPUR. 21 SHUB),' “ Alusharshid, king of Kishshatu, presented (it) to Bél from the spoil of Elam, when he had subjugated Hlam and Bara’se.” The inscription is of historical importance. We learn from it, that King Alu- sharshid subdued Elam and the country of Bara’se, doubtless in close proximity to it,” and that in the booty he carried off to Babylonia a number of costly marble vases. Part of them he dedicated to Bél of Nippur, and part, perhaps, to Shamash of Sippara, * after first having engraved upon most* of them in beautiful clear-cut characters his name and the occasion of the gift. The inscription suffices to show that Alusharshid was a mighty ruler, who in courage and adyenturous spirit was not second to Narim-Sin. But it also offers most welcome material for deter- mining the extent of the dominion of the oldest Semitic rulers. It furnishes addi- tional support to Tiele’s view (Gesch., p. 114), and at the same time proves that Winckler’s conception of the beginning of the North Babylonian history and of the extent of Sargon’s empire (Gesch., p. 38) is incorrect. Winckler proceeds upon the erroneous supposition that the deeds of Sargon, as reported in the tablet of omens and in the “legend,” are purely legendary. Hommel also (Gresch., p. 306 seq.) is ham- pered by similar prejudices. That Naram-Sin was in the possession of South Baby- lonia is demonstrated by his building in Nippur (bdni bit Bél), and by his vase found in Tello, and is furthermore established beyond all doubt by his successful operations in Magan,’ which, according to Winckler, was situated on the eastern boundary of Arabia. A vase of the Semitic king of Guti,° belonging to this same ancient period, which was probably carried by a victorious Babylonian king as trophy to Sippara, points to the extension of the power of the oldest North Babylonian rulers descriptions of campaigns in Sumerian, it cannot be surprising that the word does not occur otherwise in Sumerian inscriptions, which deal mostly with religious affairs and accounts of buildings. In favor of a Semitic etymology, to which I incline, it may be said: (1) That the word ‘‘looks very much like an original m-formation of a root ]12”’ (Jensen) and (2) that it is twice found in the Semitic inscriptions of the oldest North Babylonian rulers. 1 Tt is not to be read a-mu-ru and to be derived from amdaru with the meaning of ‘‘ersehen’’ (Hommel, Gesch., p. 302), @.¢., ‘‘to dedicate’? (Pinches, Trans. S. B. A. VIII, p. 350). Cf. Amiaud, Z A. IL, p. 296, and Jensen in Schrader’s K. B. III, Part 1, p. 26, note *°. For shub = nadanu = nadu (9), cf. 773, “gift,” Ezek. xvi. 33), cf. Tall- quist, Babylonische Schenkungsbriefe, p. 9. 2 Nothing more definite can be said at present. It is, perhaps, to be read Pura’se. Cf. the name of the mountain Ba-ti-ir (stéle de Zohab I, col. I, 7), which Scheil (l. ¢., p. 104) correctly identified with the mountain Pad(d)ir (Shamshi-Ramman Il, col. Il, 7). 3 According to Pinches Jensen, inscriptions of Alusharshid have also been found in Sippara. Cf. The Academy, September 5, 1891, p. 199, P. 8S. +A number of vases of the same high workmanship and found among them were without inscriptions. Cf. below, p. 30. 5]. R. 3, No. VII, 1. 7, namrak Magan, “plunder of Magan.” ® Cf. p. 12 seq. 22 OLD BABYLONIAN INSCRIPTIONS further northward. The inscriptions of Alusharshid testify to his supremacy over the South,’ and to his victories in the Hast and North-Hast of Babylonia. In view of all this, I regard it as impossible to question the historical character of the statements of the tablet of omens relative to Naram-Sin. Since we know that about that time a Semitie population dwelt in the northern and northeastern countries of Guti and Lulubi,” whose kings wrote inscriptions on rocks and vases in a dialect entirely identical with the Babylonian, it can no longer seem strange that Naram-Sin took the Semitic king Rish-Rammdn, of Apirak, prisoner. It is evident, however, that Apirak, which by its termination forcibly recalls names like A(E)shnunak,’ is to be sought in the North-Hast* of Babylonia rather than in the South.° If the credibility of the tablet of omens is therefore established as far as Naraém-Sin is concerned, we are no longer at liberty to call in question what it relates concerning Sargon I, unless more solid objections than have heretofore been raised, be brought against it. With Tiele, therefore, I regard as facts what Winckler describes as fiction, viz., that Sar- gon I subjugated nearly the whole world known to him, or in other words, “the four quarters of the earth.” ° But how is it that whilst Sargon always bears the title sharru dannu shar Agade or dannu shar Agade or only shar Agade,' both in the legend and in his own inscrip- 1Tncluding Lagash. Cf. p. 19. 2 This fact argues in favor of a migration of the Semites into Babylonia from the North. Cf. the ‘‘legend of Sar- gon,”’ according to which his uncle dwelt in the mountains, and he himself was carried down the river in an ark made of reed. Cf. also Winckler, Gesch., p. 141. ’ Pognon found there Semitic inscriptions written by patesis of Ashnunak. Nothing can be said with certainty as-to the exact date of these texts, but they seem to belong to the second millennium B.C. Cf. Pognon, Quelques rots du pays @ Achnounnak, read at the Académie des inscriptions et belles lettres, March 18, 1892. On this country see fur- ther Delitzsch, Paradies, p. 230 seq.; Kossier, p.60; and also Jensen in Schrader’s A. B., Part I, p. 137, note’?. * Hommel is on the right track (Gesch., p. 310, note 1). Hisreading A-ma-rak, however, has neither support nor probability. 5 Delitzsch, Paradies, p. 231, ‘‘ziemlich stidlich zu suchen.” ®Tregard also Sargon’s campaign in the West, to the Mediterranean Sea and to Cyprus, as historic facts. The cylinder of Narim-Sin’s servant found at Cyprus, and now in the Metropolitan Museum of New York (cf. Sayce, Trans. 8. B. A. VY, p. 441 seg.), has, however, no direct bearing upon the whole question. Through the kindness of Prof. Isaac Hall, Curator of the Museum, I obtained an accurate impression of the cylinder, to which, for paleographic reasons (observe, é. g., the form of the character 7a), I cannot assign an earlier date than c. 2000-1500 B. C. The pictures on it also point to a more recent date. But the cylinder is undoubtedly no modern forgery (Hommel, J. c., p. 309). 7 Nabfina’id calls him, for apparent reasons, shar Babili. It is in itself not impossible that there were kings of Babylon at some time in that ancient period. For the place where the vase of Naram-Sin was found by the French expedition, the tablet of omens (I, 7-11, cf. my restoration of this passage below, p. 26) and the occasional mentioning of Babylon (under another name) in the Sumerian inscriptions of the kings and patesis of Shirpurla clearly show that Babylon not only existed at this early time and belonged to Sargon’s kingdom, but that it even had already obtained considerable prominence (cf. below, p. 26). Cf. however, Winckler, Unters., p. 76 seg., and Lehmann, Shamashshum- ukin, p. 96, note 4. CHIEFLY FROM NIPPUR. 23 tions, his immediate successor, Narim-Sin, styles himself shar kibrat arbw2, and Alusharshid and MA-AN-ISH-TU-SU! even shar Kishshatu? This question is closely connected with the other, What do the last two titles mean? It is impossible for me to enter here into as full a discussion of this question as its importance de- mands. I therefore content myself for the present with giving the results of my investigations. As I am now considering the meaning of these titles in the earliest times only, I naturally exclude their use with the later Babylonian and with the Assyrian kings.” J. As to the Old Babylonian title, shar A’dshshatu, we have been accustomed to follow Winckler, * and to regard it as simply the equivalent of the later shar kish- shati, “king of the world.”* This identification, however, is not proved. On the other hand, it is worthy of note, (1) that supposing Alusharshid lived after NarAm- Sin, and even supposing further that he founded a new dynasty, it would still be matter for astonishment that he should exchange a title, that was not only satisfactory to Naram-Sin, known as a great conqueror, but was in itself sufficiently significant, for the synonymous shar kishshati, “king of the world;”’ (2) that no later Baby- lonian king, before Merodachbaladan I, not even the powerful Hammurabi, bears this title, though many of them apply to themselves the title shar kibrat arbwi ; (3) that Winckler’s theory, which sees in Harran the original seat of the sharrit kishshati, is improbable for the later Babylono-Assyrian time, and altogether out of question for 1 Winckler, A. K., No. 67. Paleographic reasons, the Semitic language of the inscription and the title shar Kish- shatu, establish for this king a date not only earlier than 2000 B. C. (Winckler, Gesch., p. 155), but even earlier than 3000 B. C. He is to be classed with Alusharshid. The white marble duck (Norris, On the Assyrian and Babylonian Weights, Pl. 2, No. 2), bearing the name of Vabi-shum-libur shar Kishshatu, remains without consideration here, as I do not feel at liberty to base any paleographic conclusions on the cuneiform text as it is published there. “I hope to treat’ the whole question in another place. That we may understand correctly the meaning of this title in Assyrian, the following points must be examined more carefully : (1) Is the title simply to be regarded as bor- rowed from Babylonia (cf. patesi, temple names, etc.) and extended to cover Assyrian conditions, so that only the name is Babylonian, while its semasiological development is essentially Assyrian? (2) Or, in using the title, did the Assyrians claim the same right over the same district as the Babylonians, 7. e., suppose that in Babylonia a claim was thereby expressed to Harran (Winckler), did the Assyrians by their use of the phrase make exactly the same claim upon this city? (3) Or is there no connection between the Assyrian and the Babylonian title? These questions have hitherto not been answered sufficiently. 3 Mitteilungen des Akademisch-Orientalischen Vereins zu Berlin I, p. 14. *Cf. Jensen in Schrader’s A. B. ILI, Part 1, p. 196, note 4. °If we may draw any conclusion from the later customs of Babylonian and Assyrian kings, we rather expect that in the above given case, Alusharshid, whose empire was scarcely smaller than that of Narim-Sin, according to our present knowledge, would have been particularly anxious to adhere to a title which was connected by the Baby- lonian people with the name of a very powerful ruler, and regarded by the later kings as especially important. And vice versa, if Alusharshid lived before Sargon and had founded a sharritt kishshati, “kingdom of the world,”’ it would * be strange that Narém-Sin should have used shar kibrat arba’? instead, if the other title meant exactly the same. 24 OLD BABYLONIAN INSCRIPTIONS the earliest period.’ I therefore would propose another explanation of the title, viz., to regard shar Kishshatu (or shar Kish) as identical with shar Kish, “king of Kish.”2 In other words, I infer from this title that there was a kingdom of the city of Kish similar to those of Shirpurla, Agade, ete., at the earliest time of the Baby- lonian history. Two of its rulers are so far known; both wrote Semitic, and one of them at least possessed South Babylonia and defeated Elam. Whether these kings lived after the dynasty of Sargon, or whether they preceded it and were dethroned by Sargon, will be considered below. At all events, it will be well to separate the kings of Kish* from those of Agade. There is much in favor of the view that even in the Assyrian mind‘ the title shar kishshati was originally connected with the possession of Kish, where Tiglath-Pileser III offered sacrifices to the gods (II R. 67, 11). II. But what does shar kibrat arba’i mean in the oldest Babylonian history ? After Sargon had subjugated the Elamites,’ thus fixing the natural eastern boundary of his projected great empire, he marched to the West, “ subdued ‘the land of the West,’ conquered the four quarters of the world.” The last part of the previous sen- tence, literally quoted from the tablet of omens, can in itself be imterpreted as meaning (q@) that “the four quarters of the world” lay still beyond “the land of the West,” and therefore were geographically distinct from it, or (0) that the conquest 10f. also A. Mez, Geschichte der Stadt Harran in Mesopotamien, p. 27. 2 As I remarked above, I cannot state all the reasons for my theory here. At present it may suffice to give the fol- lowing : (1) Cf. my restoration of IV R. 34, 7-11 below. (2) Cf. Delitzsch, Paradies, p.218 seg., where it is stated that the Semitic Babylonians and Assyrians wrote this city also Av-shu (and KA7i-e-ish, Brit. Mus., 82-8-16, 1, col. I, 44, pub- lished by S. A. Smith, Miscellaneous Assyrian Texts, Pl. 26; cf. also the present volume, PI. 8, No. 14, 1. 7), and Kish- sha-tu, ‘“‘according to a small unpublished vocabulary’ (ef. Paradies, p. 230). (3) Cf. also the name of the ancient king, Abil-Kish*i, known from the fragment of a Babylonian chronicle (Zrans. 8. B. A. Ul, 372), and to whom Delitzsch (Gesch., p. 72) correctly assigns the fourth millennium. 37 afterwards found that Jensen (Schrader’s K. B. III, Part 1, p. 202, note), independently of me, translated ‘‘ king of Kish” in the inscription of Manishtusu (Winckler, A. ., No. 67). His reasons for so doing and his conclusions are both unknown to me. +The facts that Ramman-nirari, who defeated the Babylonian king, Nazi-Maruttash, near Kar-Ishtar, is the first Assyrian ruler who bears the title shar kishshaté (Gin the inscription of his son, Shalmaneser I, [I R. 6, No. IV, 1. 2); and further, that Tukulti-Ninib I, his grandson, who also claims the title, must have been in the possession of Kish, as he had captured even Babylon (R. P.*, Vol. V, p. 111, col. IV, 2 seg.); and last, that neither Ashurdan I, nor Mutakkil-Nusku, nor even Ashur-résh-ishi has this title (III R. 3, No. 6, 1. 1 and 8), deserve especial attention in con- nection with my hypothesis. Afterwards the ancient meaning of the title was lost, and shar Kishshati, ‘king of Kish,’ became shar kishshati, ‘‘king of the world ’’ (which may, however, have been the very first meaning of the title before it was connected with Kish ; cf. the development of the meaning shar kibrat arba’t). 5TV R.2.34, col I, 1-3. I regard the arrangement of the individual deeds, related in the tablet of omens, as chron- ological. Among other reasons the account of Sargon’s three expeditions against the West favors this view. It was also natural that the king, before marching to the West, should protect himself in the rear by subjugating the Elamites in the East, so that during his long absence no danger might threaten Babylonia from that quarter. CHIEFLY FROM NIPPUR. 25 of “the four quarters of the world” is identical with his conquest of “the land of the West,” or (c) that the conquest of “the four quarters of the world” followed as a result upon his subduing the West. In opposition to the first view is the fact that a kingdom of “the four quarters of the world” in the far West is nowhere else mentioned, that the phrase stands without the usual determinative mtu, dlu, etc., and that this title was claimed by Babylonian kings even when they made no con- quests in the West.’ The identification of the “four quarters of the world” with “the land of the West” needs no refutation, as it has never been advanced, and in fact has no support. We can, therefore, only regard the conquest of “the four quarters of the world ” as the result of Sargon’s victories in the West, so that by the use of the title the claim is made to a quasi-worldwide dominion, as has been cor- rectly stated by Lehmann (J. c., p. 94). And indeed, Sargon, after having conquered the West, was fully justified in the Babylonian sense of the word “ world,” in thus designating his large dominion. For, in order to subjugate the West, he was obliged, because of the Arabian desert, to march victoriously first to the North, then to the West and finally southward. The enemies in the East having been previously sub- dued, and South Babylonia being also brought under his sceptre,’ he could indeed call a kingdom his own which was enclosed on all sides by natural boundaries.’ The city which had obtained the hegemony through Sargon’s deeds was A gade.’ For he calls it “my city ” (“ Legend,” 1. 26). It is the city in which he was shut up during the insurrection against him (IV _ R.’, 34, col. I, 37). And furthermore, in all his inscriptions as yet found, he calls himself “king of Agade.” But, if I understand the tablet of omens correctly, Agade does not appear to have been the capital of the empire of the four quarters of the world, as one would naturally have supposed. After Sargon had subjugated “the whole world,” he regarded as his next work the building of a capital worthy of this grand empire. The account of this important work is evidently related in IV R.’, 34, 1. 7-10, a passage’ unfortunately much mutilated and heretofore entirely misunderstood. After a careful comparison 1 Against Tiele, Gesch., p. 78. 2Tiele (J. c., pp. 738, 78) concedes the possibility, indeed even the probability of this explanation, but adds, that the title may also have had an entirely different meaning (p. 73). But what else could it have meant with Sargon I? 3This is evident from his building in Nippur, and from the fact that even his son, who was less prominent than his father, extended his influence to Shirpurla. Cf. also the express statements of the ‘‘ Legend.” 4The Elamite mountains on the east, the mountains of Armenia on the north, the Mediterranean Sea (and Cyprus) on the west and the Persian Gulf on the south. ; 5In spite of all that has been said in support of Agane, I regard this reading as improbable (cf. my remarks on Gande, p. 28). Lehmann’s statements (J. c., p. 78) prove nothing against Agade. More as to this in another place. 6 For recent translations cf. Hommel, Gesch., p. 305, and Winckler in Schrader’s A. B. ILI, Part 1, p. 102 seq. AP. S——VOL. XVII. D: 26 OLD BABYLONIAN INSCRIPTIONS of the text as given in the first and second editions of IV R.,' I transliterate and restore the passage as follows: Shar-ge-na sha ina SHIR an-ni-i Kish-shu ["?] Babilu™ t[shi-]? shum-ma epré sha “ shal-la bibu TU-N A” is-su-hu-ma .. . [ina lime?)-tu A-ga-de“ dlu i-bu-shu-ma [U B-DA]?-" shum-shii im-bu-w.... [ina lb-] bi u-she-shi-bu, “Sargon, who under this omen brought sorrow upon Kish and Babylon, tore away the earth of .... and built a city in the vicinity of (or “after the pattern of”?) Agade, called its name ‘place (city) of the world, and caused the inhabitants of Kish and Babylon (?) to dwell there.” I infer from this (a) that Kish and Babylon existed as prominent cities already in the time of Sargon I, as this great ruler deemed it necessary to render them harm- less; (0) that the dynasty of Kish was overthrown by Sargon I,° and that therefore Alusharshid and Manishtusu are to be placed before Sargon I;* (c) that the reason why the vases of Alusharshid, all badly broken, were found lying close by the com- paratively well-preserved monuments of Sargon, but not by those of Naram-Sin, is that Alusharshid apparently ruled before Sargon, not after NarAm-Sin. The question arises, Which city corresponds in later times to that built by Sargon ‘in the vicinity (?) of Agade,” and with which the title “king of the four quarters of the world”* was associated? There are reasons for identifying it with Kutha, as Winckler® does. But stronger arguments seem to point to Ursagkalama” with its famous temple, “the mountain of the world,” (always mentioned in close connection with Kish, the probable seat of the sharrit kishshati), as being identical with “the city of the world” founded by Sargon I. ‘This important text seems to have suffered still more since its first publication by George Smith in IV R.}, as a comparison with Pinches’ new edition clearly shows. Had all the differences between the first and second editions of the text, brought about by a decomposition of the tablet, been carefully noted, it would have been of great value, as the first edition is not always accessible to students. *Cf. V R. 12, No. 6, 50; II R. 52, 67 c: Ki-shu (cf. above, p. 24, note 2). Perhaps #7 is wanting, and w, ‘‘and,’’ is to be substituted. ° This is the most probable reading, according to the traces in IV R.”. Cf. K. 3657, col. I, 9 (é-shai-ush), and 1V R.? 1,* 42, a, ‘the sickness which brings woe upon the country’ (7-ash-sha-shi). *These five characters are not quite clear to me, though it is evident that Sargon purposely destroyed something. 5The two wedges beginning the character UB are clearly to be seen in IV R.!, and the last two wedges of DA still remain in IV R.*. More than two characters cannot have stood there. For the meaning of UB-DA, without arba’i, ef. Jensen, Kosmologie, p. 167. ® For various other reasons the city kingdom of Kish cannot be placed after Sargon [. 7Paleographical reasons also favor this chronological arrangement of the two dynasties. I reached my conclusion after the plates in question were printed. Pl. 4-5 and III-V are to be placed before those of Sargon I and Narém-Sin. ®It is quite possible that monuments of Sargon may yet be found, on which he calls himself ‘king of the four quarters of the earth.” %e. g., Gesch., pp. 31, 33. 10 For this reading cf. Jensen in Schrader’s A. B. ILL, Part 1, p. 22, note 5. "Cf, Winckler’s remarks, J. ¢., p. 38, in connection with ‘‘ Charsagkalama.’’ CHIEFLY FROM NIPPUR. 27 THE DYNASTY OF ISIN. Three kings of this dynasty were among the builders of the temple at Nippur, Ur-Ninib, Bur-Sin I, and Ishme-Dagdn.” Specimens of brick legends of the latter will be given in the second half of this volume. The fragment of a stone pub- lished on Pl. 9, No. 17, is unfortunately so small that we learn nothing new from it. More important are the inscriptions of both the other rulers, Pl. 10 and 11. They are taken from bricks which, at the time of their excavation, were out of their original place. These formed rather part of a platform of the Ziqqurratu con- structed or restored by Mili-Shikhu, who took them from the ruined walls of his predecessors, as old but still serviceable material for his own work. Various bricks of Ur-Ninib have thus been preserved, all with the same inscribed (not stamped) legend. Of Bur-Sin, on the other hand, only a single brick, broken in two pieces, has as yet been found. Ur-Ninib, “ Man (servant) of God Ninib,” is the king hitherto wrongly tran- scribed as’ Gamil-Ninib.* His inscription, here published, is identical with IV R2 35, No. 5. The fragment of a brick from Nippur, I R. 5, No. XXIV, erroneously ascribed to Ishme-Dag‘’n, is obviously the lower half of the same legend. In addition to the complete name of the ruler, the new text offers the correct reading of 1. 4, na-gid,* 2. e., nakidu, Hebr. 433, “shepherd ” (of Ur), and of 1. 6, mé-shi-il, “ he who delivers the commands ” (of Hridu). Bur-Sin I, so designated by me to distinguish him from another king of the same name,’ Bur-Sin II of the second dynasty of Ur,’ is a new king of the dynasty of Isin. The phraseology of his inscription is very similar to that of Ur-Ninib and Libit-Anunit’ (I R.5, No. XVIII), and thereby assures the correct reading of several characters of the latter inscription. The first sign of 1. 4 is not da (Winckler) but ingar® (identical with Briinnow, J. c. 1024), and the second sign in]. 8 is probably 1 Not Wisin, as has been generally read—last by Delitzsch, Geschichte Babyloniens und Assyriens, p. 79. Cf. the hymn 80, 7-19, 126, 1. 8, 4, published by Bezold in Z. A. IV, p. 430. 2 Pl. 9, No. 17, has been placed before Plates 10 and 11 only to save space. Ishme-Dagan was the last king of the dynasty of Isin. 3 Cf. Hilprecht in Z. A. VII, p. 315, note 1. 4 For this Semitic loan word of the Sumerian language, found also in the inscriptions of Gudea (F. col. IV, 12), cf. Jensen-Zimmern in Z. A. III, 200, 208 seg. Cf. also Jensen in K, B. III, Part 1, p. 4. 5 Although always written with the other sign Bur (Briinnow, 1. c., 9068). 6 Cf. Plates 12, 13, and Vol. I, Part 2. 7 According to Winckler in Schrader’s K. B. III, Part 1, p. 86, Libit-Ishtar. 8 Of. Jensen-Zimmern, Z. A. III, p. 199 seg. 28 OLD BABYLONIAN INSCRIPTIONS mi, not ash. lL. 3-7 in the inscription of Bur-Sin I are of special interest. They read: 3. ingar lig(?)'-ga 4. Urum™-ma_ 5. gish-kin Urudug"-ga ki-bi-gi 6. in mt-a- tum-ma~ 7. Uruy'-ga, “the powerful shepherd® of Ur, the restorer of the oracle tree* of Eridu, the lord who delivers the commands of Erech.” GANDE AND THE CASSITE DYNASTY. A number of inscribed objects excavated in Nippur bear the name of a king’ who has been transliterated Glar-de (?) by Pinches.° As I remarked in another place,’ this transliteration is incorrect. For the first character of the group on PI. 14, No. 23, 1. 2b, is not the Old Babylonian sign for GAR,’ but GAN.’ The second character may be read either de or ne, the whole name therefore, either Glande or Ganne. The former reading is the more probable, because the second character, out- side of the purely Sumerian” texts, is more frequently found with the syllabic value de than ne." The contents of the three inscriptions of Gande published on PI. 14 are iden- tical. They read: 1. ?"!Hn-lil-la 2. lugai ki-aga-ni Glan-de 3. a-mu-na-shub, “To 1 Cf. Jensen, Z. A. I, p. 396, note 4. > mi-a-tum-ma, corresponding to mi-shu-il (Ur-Ninib, 1. 6), as tum, like il, is explained by abdlu, ‘‘to bring, to deliver.”’ Cf. IV R.? 35, No. 6, 12, 13. ° Of. tk-ka-ri Ba-bi-i-lu ‘“‘(Nebuchadrezzar II), shepherd of Babylon’? (Abel-Winckler, Keilschrifttexte, p. 33, 1. 19). Ingar = tkkaru, Hebrew 128, is a Semitic word adopted by the Sumerian language (Zimmern, Babylonische Busspsalmen, p. 5, note 1), and means ‘“‘ farmer,’ Landmann (Jensen-Zimmern, in Z. A. III, p. 199 seg. ; Delitzsch, Assyrisches Worterbuch, pp. 400-402). In view of the principal occupations of the farmer—tilling of the ground and stock-raising—the word occurs as a synonym either of irrishu, talm. NDS (Z. A. III, p. 200), or of nakidu, rid alpi (Z. A., ibid.). Accordingly, it is to be translated either as “‘farmer”’ or as ‘‘shepherd.’’ The latter meaning is the only possible one in the above-given passage, as the context and a comparison with Ur-Ninib, 1. 4—na-gid Urum*- ma, ‘shepherd of Ur’—clearly show. The same meaning is also to be preferred to Landmann (Jensen, in Schrader’s K. B. Ill, Part 1, p. 59) in passages like Gudea F, col. III, 1. 14, where ingar stands parallel with utul, sib and nagid, all words for ‘‘shepherd.”’ * Cf. Jensen, Kosmologie, pp. 99 seqg., 249, note. ° That the bearer of this name was a king is certain (against Pinches), notwithstanding the omission of the title. Cf. Hilprecht, “‘ Die Ergiinzung der Namen zweier Kassitenkénige,’”’ Z. A. VIII (in print). ® The Academy, 1891, September 5, p. 199, a, b. ‘Z. A, VII, p. 315, note 1. ® Amiaud et Méchineau, U. c., No. 105. ® Ibidem, No. 79, sign 5. 10 To be understood in the sense established by Lehmann, Shamashshumukin, pp. 62-108. 1 For this and other reasons I reject the reading Agane instead of Agade (= Akkad/ in spite of Lehmann, Sha- mashshumukin, p. 73). Cf. also Hommel, Gesch., p. 302. CHIEFLY FROM NIPPUR. 29 Bél, his beloved lord, Gande has presented it.” But who was this Gande who left his name on a number of marble vases,’ on a large unhewn block of white marble, on two others of reddish granite and on the edge of two door sockets belonging to former Babylonian kings? A due consideration of the following points will enable us to answer the question. 1. The short inscription of Gande just translated is written not only on his own monuments by this king, but is also found on the rough edges of a door socket of Sargon I, and another of Bur-Sin IT. Hence it follows, that Gande must have lived after their time, 2. ¢., after c. 2400 B. C. 2. On the other hand, it follows from the depth of the place in which the stones were found and also from the peculiar characters of the inscriptions (see below), that Gande could not have ruled after Mili-Shikhu, or, as the immediate seven or eight predecessors of the latter are known, not after c. 1240 B. C. 3. It is remarkable that Gande by two of his inscriptions characterizes door sockets which had previously been presented to the temple as his own gifts. It is in itself clear that these inscriptions cannot be regarded in the sense of inventory labels, as they are sometimes found in connection with Egyptian antiquities. Only one explanation seems possible, namely, that Gande was not a native king, but invaded and conquered Babylonia and regarded the property of the temple in Nippur as his legitimate spoil. As however he, with his victorious hordes, did not leave the subjected country again, but usurped the Babylonian throne, thereby becoming the founder of a new dynasty, the conquered cities and temples became part of his new empire, to which he now restored the trophies of his victory as his own personal gifts. Had he left Babylonia, he certainly would have carried away the treasures of the temple as spoil to his own country, just as Alusharshid and Naram-Sin did, after they had conquered Elam and Magan, or Nebuchadrezzar 1, atter the destruction of Jerusalem. 4, This explanation of Gande is supported by the character of his inscribed objects and by the peculiarity of their cuneiform writing. All his inscriptions are carelessly executed and are engraved very shallowly ; indeed, those on the door sockets and large blocks are only scratched in the unhewn stone. Besides, the char- acters employed violate the laws which underlie the regular development of the Babylonian cuneiform writing. They appear to have been cut by men unaccustomed to use the chisel in writing, who, it is plain, had adopted the Babylonian system of writing, even endeavoring to imitate the characters of a certain period,? but who were neither familiar with their original meaning, nor with the 1Cf. Vol. I, Part 2. 2 Cf. e. g the characters of the inscriptions of Ur-Nina, de Sarzec, Décowvertes, Pl. 31, No. 1. 30 OLD BABYLONIAN INSCRIPTIONS exact form then in use. The scribe regarded e.g. GAN (PI. 14, No. 23) as the doubled form of a certain sign resembling the reversed ancient SAG.’ For occasion- ally he divides this character into halves, placing one after the other (Pl. 14, No. 24, 25). The artistic execution of the vases themselves stands in striking contrast to the rude appearance of the inscriptions on them and on the large stones. As a num- ber of uninscribed vases of similar form and of the same skillful workmanship were found together with those of Alusharshid, there is every reason to believe that Gande’s vases formed originally part of the former’s gift to the temple, the more so as they were found in close proximity to those of that very ancient king. Only the unhewn blocks of marble and granite, apparently intended for door sockets, were genuine gifts of Gande, probably brought from the Elamite mountains. From the fact that the place occupied by the inscription was not polished or even smoothed, we likewise infer that the scribes of this ruler had neither the artistic taste nor tech- nical training of the Babylonian stonecutters. 5. The name Gande has not a Babylonian sound. Besides, it is sometimes found abbreviated into Gan. This peculiarity of abbreviating names is characteris- tic of the rulers of the second and third dynasties of Babylon, as is shown by com- paring List b with List a and with the inscriptions of Bibeiashu.” Only one king fulfills the requirements (viz., a foreigner, founder of a new dynasty, a prince whose name begins with Gan, and who lived between ¢. 2400 and ¢. 1240 B. C.). This is Gandash, the first ruler of the Cassite dynasty, which occupied the throne of Baby- lonia for five hundred and seventy-six years. Gande (otherw. Gan) is abbreviated from Gandash* in the same way as Bibe from Bibeiashu.’ It is significant that, with the exception of fragment Brit. Mus. 84-2-11, 178 (see note 3), no monument of the founder of the Cassite dynasty and very few of its other members have, up to the present, been found outside of Nippur. This latter was, as I shall later show in detail, the very centre and stronghold of the Cassite dynasty. It is not, therefore, accidental, that the representatives of this foreign house dedicated so many valuable gifts to the temple of Bél in Nippur. By not paying the same hom- age to Marduk of Babylon and his illustrious city, which Hammurabi’ had endeavored to raise to the most prominent position in the political and religious life of the country, 1 Amiaud et Méchineau, J. c., No. 221. 2 Cf. above, p. 17. ’ Who again is identical with the Gaddash of Brit. Mus. 84-2-11, 178 (Winckler, Unters., p. 156, No. 6). Cf. Hilprecht, Z A. VII, p. 309 seg., especially note 4. * Cf, Hilprecht, ‘‘ Die Ergiinzung der Namen zweier Kassitenkénige”’ in Z. A. VIII (in print). ° It is worthy of notice, that not one votive object with an inscription of a ruler of the first or second dynasty of Babylon has so far been found in Nippur. These kings concentrated their attention on the glorification of Babylon. CHIEFLY FROM NIPPUR. 31 but by restoring the former glory of Ekur, the ancient national sanctuary in Nippur, so deeply rooted in the hearts of the Babylonian people, and by stepping forward as the champions of the sacred rights of “the father of the gods,” they were able to bring about a reconciliation and a final melting together of the Cassite and Semitic elements. Supported by the influential priesthood of Nippur and dreaded as daring warriors by the discontented parties, the Cassites could mould and govern the desti- nies of Babylonia for nearly six hundred years, until finally they were overwhelmed by new invasions from the East and by the great national uprising in the South, which resulted in placing the native dynasty of Pashe on the throne of Babylon. The essential results to be drawn from the fifty-five votive inscriptions of the Cassite dynasty published on Plates 14-29, I have given in several articles in Zeztschrift Sir Assyriologie* and may therefore confine myself to the following points. The inscriptions on Pl. 8, No. 15, and PL.-21, No. 43, are written on the obverse and reverse of a tablet in agate. The stone tells its own story. About 2750 B. C., the patesi ® of a city dedicated the tablet to the goddess Ninna or Ishtar “ for the life of Dungi, the powerful champion, king of Ur.” Afterwards, possibly about 2285 B. C., at the time of the Elamite invasion, when Kudur-Nankhundi laid hand on the temples of Akkad and carried the image of the goddess Nanf into Elam, the tablet was also taken away and remained in the possession of the enemies until c. 1300 B.C. Kurigal- zu (doubtless the second of the name“), after his conquest of Susa, brought it back to Babylonia and presented it to Béltis of Nippur. For over three thousand years it lay within the walls of Ekur, until again it became the spoil of invaders of Nippur. This time it was carried far away to the modern" Aharri. Perhaps a later shar kibrat arbwim will take it back to the resurrected sanctuary of Nippur. Kuri- galzu’s inscription on this tablet is of bistorical importance, because, for the first time, we learn from this king’s own inscriptions of his successful campaign against Elam,' in the course of which he conquered even Susa.” The cuneiform text reads: 1. Kurigalzu 2. shar Karuduniash 3. ékalla sha“ Shdsha™ 4. sha Elamti™ 5. ikshud- ma 6. ana Bélit (NIN-LIL) T. béltishu 8. ana baldateshu 9. tkish, “ Kurigalzu, king of Karuduniash, conquered the palace of Susa in Elam and presented (this tablet) to Bélit, his mistress, for his life.” 1 Inscription of Kadashman-Turgu, Pl. 24, No. 63, 1. 1 and 2. 2 Cf. “ Bibliography,’’ II, 9, 11, 12. 3 This word stood apparently in one of the lost lines at the lower end of the tablet. 4 Of, Pinches, ‘An Early Tablet of the Babylonian Chronicle,” in R. P.?, Vol. V, p. 109, col. IIT, 10-18. 5 The earliest mention of Susa in the Babylonian cuneiform literature. The absolute proof for the identity of Shasha with Shishi (LV R252, 46, bs IL R. 48, 59, 6, and Delitzsch, Paradies, p. 326), Shishan or Shushun, is impossi- ble at present. It seems, however, scarcely possible that ékallu sha Shasha sha Hlamti can be anything else than pyya we TYAN ww (Dan. viii. 2). The name was probably pronounced Shosha(n). Cf. also p. 13, note 1 (end). 32 OLD BABYLONIAN INSCRIPTIONS Another inscription published on the same plate, Nos. 41 and 46, was damaged at the end of each line when the scribe cut it from the block of lapis’ lazuli,* which Kurigalzu dedicated to Bél. It reads: 1. A-na Bel (Hn|-lil] ) 2. be-el ma-ti-a-te be- [Nishi] 3. Ku-[r]i-gal-zu ri-ia-um [na-ram “ Bélit ? |’ 4. pa-li-ch | she-mu-u i Bél?],° “To Bél, lord of the lands, his lord, Kurigalzu, the shepherd beloved by Bélit, he who fears (and) obeys Bél.” The cuneiform text of the lapis lazuli disc on Pl. 23, No. 61, proves the correct- ness of my conjecture in Z A. VII, pp. 305-318. The fourth character of |. 3 is, however, not as I supposed, Ka but Aad.‘ The disc thus furnishes us the new and interesting writing kaddashman”® instead of the hitherto hadashman. No. 66 and 67 of Pl. 25 are the obverse and reverse of the same fragment of an agatering. The dedication onit was apparently written by one king only, who, in need of space, inscribed both the upper and lower side of his gift. As the remnant of the last character of No. 66 is doubtless to be completed to Ka-[dingir-ra™], the ideo- gram shar, standing before it, must be the title of a king, whose name ended in LIL (the last character of “"""HN-LIL or Bél). According to our present knowledge of the rulers of the Cassite dynasty, the name can be read either Audur-""" HN- LIL’ (cf. No. 64) or Kadashman-""""EN-LIL (No. 65). The obverse of the ring (No. 67) contains part of a name ending in [b]u-r7-cal-ash], which again can be completed either to Shagashalti-Buriash, the son of Kudur-“""HN-LILD, or’ te .... buriash (No. 68, col. I, 5), the son of Kadashman-“"""HN-LIL. As no in- scriptions of the former seem to have been found in Nippur, and the characters of Nos. 66 and 67 resemble those of No. 68 more than of No. 64, I assign the ring to the king mentioned in No. 68, 7. e., in all probability Kadashman-Buriash, who, according to III R. 4, No. 1, was at war with an Assyrian king.® The following 1 Cf. Hilprecht, ‘“‘Zur Lapislazuli Frage im Babylonischen,”” Z. A. VIII (in print). 2 Briinnow, J. c., 5309. Cf. Meissner, Beitrage zum Altbabylonischen Privatrecht, p. 115, No. 21, 3. 3 Uncertain ; restored according to Brit. Mus., 81, 8-80, 9, 1. 8,9 (cf. Jensen, Schrader’s A. B. III, Part 1, p. 120): ri a(sic! instead of Jensen’s ’w)-u na-ram tuBelt, pal-hu she-mu-u tu. Shamash. 4 Briinnow, J. c., 2701. See also my ‘‘Nachtrag’”’ in 7 A. VII, p. 318. 5 This is not to be used in favor of Pinches’ identification of kKaddash with gaddush and gan(kan)-dash. I adhere to what I remarked in Z. A. VII, p. 309, note 4, until Gaddash or Gandash, the founder of the Cassite dynasty, has actually been found written with the character Aa (or Ka), or the word /ad(d)ash in Cassite proper names like. Kad- (d)ashman-Turgu, with the value ga (or ka). Cf. Pl. 25, No. 68, col. I, 14, 15, dwmu sag Kad-ash-ma-an-dingit HN-LIL, “(....riash) the first son of Kadashman-EN-LIL.”’ My writing dumuw Ka-da-ash-ma-an-Uingi Bel (Z. A. VII, p. 309, note 3) is to be corrected accordingly. 6 Generally read Audur-Beél. ‘It would be more appropriate to transliterate him Kudur-Turgu (see below). That he was king will be shown in my article, ‘‘ Die Ergiinzung der Namen zweier Kassitenkonige,’’ Z. A. VIII (in print). 7 For various obvious reasons other possibilities have been excluded as improbable. ® The conjecture of Delitzsch (Hossaer, pp. 10 seg.; Hommel, Gesch., p. 437 seq.), that the Assyrian king was Shalmaneser I, is proved by the new chronology which I am able to establish for a number of Cassite kings. Cf. below p. 37. CHIEFLY FROM NIPPUR. 33 is an attempt to restore the legend according to the usual phraseology of this class of inscriptions: Obverse, [°Hn-Uil lugal-a-ni-(ir) Ka-da-ash-ma-an-B \u-ri-ia- [ash], Reverse, [dumwu (sag) Ka-da-ash-ma-an-"En}-lil lugal Ka[-dingir-ra™ a-mu- na-shub], “To Bél, his lord, Kadashman-Buriash, (first) son of Kadashman-EN- LIL, king of Babylon, presented it.” The question remains to be settled, whether the name of the father of Kadash- man-Buriash is to be read Kadashman-Bél, as has generally been done,‘ or Kadash- man-Enlil’ or still in another way. The second reading necds no refutation. It is in itself impossible. The first seems to me at present improbable. For while there are Babylonian proper names which are composed of Babylonian words and the name of a foreign god,* there is no evidence that there were in use any which contain a Cassite word and at the same time the name of a Babylonian deity. The example quoted by Delitzsch' should be read Nazi-Shibhu.’ For this very reason I regard the correct pronunciation of Kadashman-""""HN-LIZ as being either Kadashman- Kharbe® or Kadashman-Turgu," in other words the Cassite king Kadashman- “a'r JN-LIL may represent either of the two persons. Which of the two is the more probable? There are two Cassites of the name Kadashman-Kharbe to be considered. The one was the father of Kurigalzu I.’ As, however, there is no proof that he was a king,* we leave him here out of consideration, the more readily, as other reasons make his identification with Kadashman-“""HN-LIL well-nigh impossible. The other Kadashman-Kharbe is entirely out of the question,’ as none of the six kings following the latter successively, according to List b, ends in . 1¢. g., Delitzsch, Kossier, p. 20 ; Pinches, The Academy, September 5, 1891, p. 199, 6, and last Hilprecht, Z. A. VIL, p. 316. 2 Hommel, Gesch., p. 433 : Kara-Inlil. 3@. g., Shuhamuna-ah iddina (Delitzsch, Kossier, pp. 18, 21, 28), Kashshit-nadin-ahu (ib.). 4 Kossaer, p. 18, note 1. 5 For Cass. Shilw = Babyl. Marduk cf. Delitzsch, Kossaer, pp. 20, 21, 39. From the few published documents in which Nazi-Shihu or members of his family (cf. the passages on p. 42) are mentioned, it is evident that this Cassite family lived in Northern Babylonia and was very prominent and influential. Hven Nebuchadrezzar I, shalilu Kash- shi, treated its chief with distinction (Hredbrief, col. II, 12: Kalu Akkad). In view of the true character (Hil- precht, Z. A. p. 311, note 3) of the so-called “ Cassito-Semitic vocabulary ”’ (Delitzsch, Kossaer, p. 24 seq.), and of what has been said about the formation of proper names above, I believe Nazi-Shihu in V R. 44, 48a, to be the same person as the high dignitary who appears as the first witness in the “ Freibrief ’’ of Nebuchadrezzar I. 6 For Kharbe = Bél cf. Delitzsch, Kossaer, p. 28; for Turgu = Bél cf. Hilprecht, Z. A. VII, p. 316, note 3, and the following lines above. 7 Cf, Winckler in Z. A. II, pp. 307-311 8 Avainst Delitzsch, Gesch. (‘‘ Ubersicht’’), who does not hesitate to number him among the Cassite rulers. °The principle stated by Winckler in Z A. IL, p. 310, 1. 7-10, is correct, but his identification of Kadashman- Bél with Kadashman-Kharbe is impossible. A. P. S.—VOL. XVIII. E. 04 OLD BABYLONIAN INSCRIPTIONS riash, as is required.! That Turgu is another Cassite equivalent for the Babylonian Bél (of Nippur), I have endeavored to show in Z. A. VII, p. 316, note 3. But there are other reasons for identifying Kadashman-Turgu with Kadashman-“"" HN-LIL - (1) The cuneiform characters of the inscriptions of Kadashman-Turgu on Plates 23, 24, are strikingly similar to those of Kadashman-“""HN-LIZ and especially his son (Pl. 25). (2) The son of Kadashman-”""HN-LILT bears precisely the same title (PI. 25, No. 68, col. I, 6), as Kadashman-Turgu (PI. 24, 1. 8).? On Pl. 28 we meet with the first personal inscription of Rammdn-shum-usur, contemporary of the Assyrian king, Bél-kudur-usur. The brick legend is written in Sumerian and reads: 1. 29" Hn-lil 2. lugal kur-kur-ra 3. lugal-a-ni-ir 4. 2!" Rammén-shum-usur 5. siba she-ga-bi 6. v-a En-lil"-a 7. sag-ush E-kur-ra 8. E-kur e ki-ag-ga-a-ni_ 9. °°" al-ur-ra-ta 10. mu-un-na-ru, “To Bél, lord of lands, his lord, Ramman shum-usur, his favorite shepherd, adorner of Nippur, chief of Ekur, built Ekur, his beloved house, with bricks.” ! Winckler, following Sayce,* latterly inclines to regard the Babylonian king “ Ramman-shum-nasir,” in III R. 4, No. 5, as identical with the ruler whose inscrip- tion has just been translated.“ This, however, is utterly impossible. Sayce and Winckler misread the name of the king mentioned in III R. According to the law underlying the formation of Babylono-Assyrian personal proper names, the cuneiform group Rammdan-MU-SHESH-IR can only be read Rammin-mushéshir, “ Ramman is directing (ruling).”°? This king lived before Burnaburiash and has not even the name in common ‘with the above-given Rammaén-shum-usur. 1 For Kadashman-dingir#N-LIL, himself king (Pl. 25, No. 65), was the father of another king (Pl. 25, No. 68, col. I, 16), endingin .. . . riash (ibid., 1. 5). * Besides the personal votive inscriptions of King Kadashman-Turgu, many tablets dated in his reign were found in Nippur. It is certain that he was one of the best known princes of the Cassite dynasty and ruled more than fifteen years. It seems, therefore, strange that his name, being entirely Cassite, should have been omitted by the compiler of K. 4426 (V R. 44, 21-44, a, b). As soon as we read the name in V R. 44, 29, a, Kadashman-Turgu, as I proposed above, the difficulty is removed. And, indeed, this reading finds new confirmation. All the names placed together by the compiler in V R. 44, 23-44, are purely Cassite. Therefore we are obliged to regard the ideogram in the name of Kadashman-dingir i N-LIL, which is explained by its Assyrian equivalent Twkulti-Bél in the right column, as Cassite in the left column. That dgir#N-LIL was not pronounced Kharbe seems, apart from the above-given reasons, to be indicated by the fact that Kharbe in V R. 44, 33 a (i.e., in the left column) is written phonetically Khar-be. From names like Kharbi-Shihu (IV R.* 34, No. 2, 1. 5, 14), ‘Bél (= the lord) is Marduk,’’ we may infer that the real mean- ing of Kharbi was something like ‘‘lord.’”? The use of Kharbi for the nameof a certain god, resembles, therefore, closely that of “ngir#'N in the later Babylonian time (cf. Tiele, Gesch., p. 538). Turgu on the other hand seems to have been the Bél of the Cassites, 7. e., exactly corresponding in his rank to the dingir#N-LIL or Bél of Nippur, the highest god of their Pantheon. 3h. P.?, Vol. Il, p. 207, note 1 (ef. Vol. I, p. 16). * Gesch., p. 102 (cf., however, pp. 88, note, and 157). ° Cf. u-shesh-she-ru, Sanh. Kuy. 2, 31. CHIEFLY FROM NIPPUR. 35 The brick legend on PI. 29 was already published by Pinches in Hebrazca, Vol. VI, pp. 55-58. I need make no apology for republishing it here, as Mr. Pinches’ edition, I am sorry to say, is of little use, the cuneiform text and translation offered by him being unfortunately incorrect in all essential points. The legend was stamped ““by means of a wooden block, on the brick.” The stamp, however, having been carved very shallowly, the inscription, “ though impressed evenly,” is not very distinct on any of the many hundreds of bricks which were found.' Besides, the surface is covered “ with a thin deposit, which adds to the difficulty of deciphering the in- scription.” Notwithstanding all this, I did not deem it necessary to mark any of its cuneiform characters as doubtful. My copy was made after a long and careful study of each character, and especial attention was paid to every detail. Certain cuneiform characters could not be recognized distinctly on the original except in the light immediately preceding sunrise, the best time for copying difficult cuneiform in- scriptions. On the following points I am obliged to differ from Mr. Pinches: 1. Pinches: “The date of this inscription is uncertain. Judging from the style of the characters, it should be about 1500 B. C., but it may be as early as 2500 B. C.” In the present writer’s opinion the inscription belongs to one of the last rulers of the Cassite dynasty. For paleographic reasons it cannot be older than 1250 B. C., and in fact belongs to a king who ruled ec. 1165 B. C. 2. Pinches transliterates the name of the ruler (1. 4) “ Nin-Dubba,” regards its bearer to be a lady, and adds, the inscription “is the only text of a queen of Meso- potamia known.” Mr. Pinches should have been the more careful in introducing this regent as a female to Assyriologists. I read 1. 4 Mili-Shikhu (see below) and regard - this person as being the well-known Cassite king who ruled ¢. 1171-1157 B. C. 3. The first character in 1. 5 is, according to Mr. Pinches, nin, “lady,” while in reality the text gives szba, “shepherd.” 4, Mr. Pinches reads (1. 6) lugal Hga, “ queen of Ega,” and adds, “ Ega is probably another name for this city [Nippur], or for a part of it.” The phrase thus misunder- stood by Mr. Pinches is the very common title lugal lig (?)° -ga, “the powerful king.” The inscription in question reads as follows: 1.?Hn-lil-la(l) 2. lugal kur- kur-ra_ 3. lugal-a-ni-ir 4. ?"" Mili-"""Shihu 5. siba she-ga-bi 6. lugal lig (?) -ga 7. lugal ub-da tab-tab-ba 8. H-kur 9. e-ki-ag-ga-a-ni “10. °al-ur-ra-ta 11. mu-un-na-ru, “To Bél, lord of lands, his lord, Mili-Shikhu, his favorite shep- herd, powerful king, king of the four quarters of the earth, built Ekur, his beloved house, with bricks.” 1Cf. ‘‘Table of Contents.’’ * Jensen in Z. A. I, p. 396, note 4. 36 OLD BABYLONIAN INSCRIPTIONS My reasons for identifying the name in |. 4 with that of Mili-Shikhu are as fol- lows: (1) The king must have lived after Ramman-shum-usur, because a few bricks of the latter’ were found in the platform of the temple erected by him.’ (2) Paleo- graphic reasons point to the end of the Cassite dynasty as the date of his inscription. Apart from a certain difference of appearance between RammAan-shum-usuwr’s legend and that of the king in question, the one having been inscribed, the other stamped, there is a decided similarity between the characters of the two inscriptions. (3) One of the titles (1. 5), the pbraseology of the beginning (1. 1-3), and—what is especially characteristic—that of the end of the two inscriptions (1. 8-11, otherw. 10), in other words, 8 (otherw. 7) lines are absolutely identical. Hence it follows that the king in question must have ruled not long after Ramman-shum-usur; was possibly his suc- cessor. (4) This result is corroborated by an analysis of the first half of 1. 4. The determinative dingir is not unfrequently found before the names of Cassite kings.’ The second and third characters are to be read SHA (libbu)'+ ba. The absence of the two inner wedges in SZZ/A is due to the shallowness with which the characters of the stamp were carved. They are found on another (badly preserved) brick, of the same king, the legend of which was written with the hand, and differs slightly in other respects.” As the inscription is written in Sumerian, the syllable ba indicates that the Sumerian value of the preceding sign ended in 0, in other words, was the dialectic form of a word ending in g—probably shag. As the personal proper names occurring in the later Sumerian inscriptions are, as a rule, not to be read Sumerian, but as they were actually pronounced,® we read the ideogram (shaba) with one of its common Semitic equivalents, kirbu, libbu, milu, ete.’ : Only one of the Semitic ideographic values of this character fulfills the require- ment of forming the beginning of one of the well-known names of the last four Cassite kings, ¢. ¢., milu or mili. As, on the other hand, there is only one Cassite king of that period who begins with Mili, I confidently believe the last group of cuneiform characters in]. 4 to be an ideogram for the god Marduk, or his Cassite equivalent Shikhu, and read the whole name accordingly Mili-Shikhu. The following list is an attempt at restoring part of the broken List b, and giv- ing the chronology and succession of the last twenty-four kings of the Cassite 1 Together with a few of Ur-Ninib, Kurigalzu, and one of Bur-Sin I. “Cf. above, p. 27, and ‘‘Table of Contents,” Pl. 29, No. 82. * Cf. Hilprecht in Z. A. VII, pp. 308-310. 4Cf. Briinnow, 1. c., 7983. 5Cf. Vol. I, Part 2. ®Cf. also Jensen in Schrader’s H. B. III, Part 1, p. 117, notes 6-9. 7Cf. Briinnow, /. c., 7985-7992. CHIEFLY FROM NIPPUR. 37 dynasty, which ruled over Babylonia for 576 years. My reasons for changing the generally accepted order of several of these kings will be found in a special article. If the date which I assigned to the first rulers of the Pashe dynasty be accepted, my chronology from Kurigalzu II to Bél-shum-iddina II must be regarded as abso- lutely certain. As the rulers between Barnaburiash and Kurigalzu IT are well known, it is also settled beyond doubt that Shagashalti-Buriash lived before Kuri- galzu I. Nabuna’id’s statements concerning the chronology of Sargon I, Hammu- rabi, Burna-Buriash, and Shagashalti-Buriash must be regarded as only approximate dates. The events recorded may have occurred at any time in the century before or after the year given.” Sennacherib’s statement concerning Tukulti-Ninib’s cylinder (600 years) is likewise to be understood in a broad sense. 13. RammAn-mushéshir®. . .. . . . c.1442-1423 (about twenty years). 14. Kallima(?)-Sin....... . . .¢, 1422-1408 (about fifteen years). WS, Ieoietnve bine 5 5 56 6 6 . . . .c. 1407-1393 (about fifteen years ?). 16. Shagashalti-Buriash (his son) . ¢. 1392-1373 (about twenty years). 17. Kurigalzu I (son of Kadash- man-Kharbe)....... . .¢c, 1372-1348 (about twenty-five years). 18. Kara-indash (his older son?)’ . c.1347-1348 (about five years?). 19. Burna-Buriash (son of 17) . . c. 1342-1318 (about twenty-five years). 20. Kara-Khardash (son of 18) . .c.1317-1308 (about ten years). 21. Nazi-bugash (usurper)®. .. . .c.1307 (about one year). 22. Kurigalzu II (son of 19)... . 1806-1284 (nearly twenty-three years). 23, Nazi-Maruttash (his son) ... 1284-1258 (twenty-six years). 24. Kadashman-Turgu (his son)’. . 1257-1241 (seventeen years). 25. Kadashman-Buriash (his son) . 1240-1239 (two years). AG, ISIN 6 65 Who pe bo oe 1238-1233 (six years). 27. Shagashalti-Shuriash*...... 1232-1220 (thirteen years). 1T regard Peiser’s doubts as to the correctness of the 576 years (Z. A. VI, p. 267 seq.) as unnecessary. Through the excavations at Nippur we are enabled to substantiate part of the statements given as to this dynasty in the list. This fact teaches us Festina lente! 2 And in a sentence like ‘‘ who built 700 years before Burnaburiash,’”” we have to make even a greater allowance, as we do not know which approximate date Nabuna’id had in mind in connection with the reign of Burnaburiash. 3 He may have lived at an earlier date. ‘Generally read Kudur-Bél. Cf. above, p. 32 seg. 5The same as Kar-indash, son-in-law of Ashur-uballit, king of Assyria. Cf. R. P.’, Vol.V, p. 107, 1. 5, 6, 12. 6 Called Su-zigash in R. P.”, Vol. V, p. 107, 1. 10, 18. TOCf, Hilprecht in Z. A. VII, p. 317 (cf. Pl. 23, No. 61). The date there assigned to Kadashman-Turgu (c. 1340 B. C.) is to be corrected according to that given above. For his identification with Kadashman-dingirHN-LIL see above, p. 33 seq. 8Cf. above, p. 11. 38 OLD BABYLONIAN INSCRIPTIONS 28. Bibe[iashu] (his son)'. ..... 1219-1211 (nine years). 29, Bél-shum-iddinaI........ 1210-1209 (one year and a half). 30. Kadashman-Kharbe...... . 1209-1208 (one year and a half ). 31. RammAan-shum-iddina. ..... 1207-1202 (six years). 32. Ramman-shum-usur....... 1201-1172 (thirty years). 33. Mili-Shikhu (his son)’. . 1171-1157 (fifteen years). 34. Marduk-abal-iddina (his Leas 1156-1144 (thirteen years). 35. Zamama-shum-iddina...... 1143 (one year). 36. Bél-shum-iddina IT*,...... 1142-1140 (three years). The last 24 kings = c. 303 years; the first 4 kings = 68 years; the remaining 8 kings = 205 years and 9 months (each 25-26 years in average’). Total, 36 kings = 576 years and nine months. THE DYNASTY OF PASHE The cuneiform tablet published on Pl. 30 and 31 forms a part of the collection J. S., purchased by the Expedition from Joseph Shemtob® for the University of Penn- sylvania, July 21, 1888. Unfortunately it is impossible to ascertain with certainty where the stone tablet was found.’ In regard to its size and mineralogical character it closely resembles the “black stone of Za’aleh,” to be found in I R. 66, with which it also has much in common as to its contents. Both belong to the class of the so- called kudurru inscriptions.’ A piece of ground situated in the land of Kaldi, in the province of Bit-Sinmagir (I, 1, 2), which for many years (I, 3-8) had been in pos- session of the family of a certain Nabfi-shum-iddina (i, 15) but had been unlawfully reduced in size by Hkarra-ikisha, at that time governor of Bit-Sinmagir (UG Bab). was upon the complaint of the owner (I, 16-II, 5) restored to its original extent by 1Tdentical with S. 2106, 1.9. See above, p. 11. 2Cf. Belser in B. A. II, p. 197, 1. 31. SCf. R. P.2, Vol. V, p. 111, 1.14; p. 112, 1. 16. Cf. also below, p. 41. 4Such long reigns appear in no way improbable when compared with the longer reigns of fifteen rulers of the first and second dynasties of Babylon. 5 Sayce (R. P.2, Vol. I, p. 17, note 8) regards this city as identical with Isin and Patesi. Cf. II R. 53, 18a. ®° Cf. Harper, Hebraica V, pp. 74-76. 7 Cf. ‘‘Table of Contents,’’ Pl. 30, 31. 8T reckon as such not only “those Babylonian documents which are inscribed on blocks of stone not always quite regularly hewn’ (Belser, B. A. II, p. 111), but also those which, like ours and the Za’aleh stone, were kept within doors and possibly as duplicates of the ‘‘stéles,’’ which were naturally exposed to destructive influences, so that in disputes concerning boundaries they might furnish the basis for a legal decision. CHIEFLY FROM NIPPUR. 39 Bél-nadin-aplu, king of Babylon, in the fourth year of his reign (II, 6-10). The document closes with a blessing for the official who in time to come shall respect the decision (II, 11-20), and with a curse against him who shall remove the boundary again (II, 21-24). Apart from the fact that the stone furnishes us with the name of one of the early kings of the “Sea-land,” with that of a hitherto unknown province or county of the land of Kaldi,' and with other details of interest, it is of the greatest importance for its chronological bearings. For the following reasons, the stone must be assigned to the Pashe dynasty: (1) The cuneiform characters are those which are characteris- tic of the documents of that period, and especially they resemble those of the charter (freibrief ) of Nebuchadrezzar I.* (2) Ekarra-ikisha, son of Ea-iddina, is mentioned as an official’ both on our stone (I, 10, 11; II, 6) and on that of Za’aleh (II, 6). From this it follows that our stone belongs to about the same time as the other which bears the date of the first year of King Marduknadinahé. (3) But we are able to fix the date of our stone even more exactly from the statement in col. I, 7-15, according to which the piece of land in question was in possession of the family of Nabfi-shum-iddina until the time of Nebuchadrezzar I, but in the fourth year of King Bélnadinaplu was unlawfully encroached upon by the governor, Ekarra-ikisha. The result naturally is that the stone dates from the reign of Bélnadinaplu, and that the latter was the immediate successor of Nebuchadrezzar I. This proves, at the same time, that the supposition made by Winckler‘ and Delitzsch,’ that Marduknadinahé was the immediate successor of Nebuchadrezzar I, is wrong, and that the order is rather Nebuchadrezzar I, Bélnadinaplu, Marduknadinahe. The question arises, What place must be assigned to this group of three kings in the dynasty of Pashe? ‘This, in my opinion, can be answered with entire certainty. For although the Babylonian list’ has been broken off at the very place where the names of the rulers of this dynasty once stood, yet the characters which remain of the last three kings serve us in solving the question. Of the five known kings of this dy- nasty, 1. Nebuchadrezzar I, 2. BélnAdinaplu, 3. Marduknadinahé, 4. Mardukshapik- zirim (ste!) (not Marduktabikzirim)’ 5. Rammanapluiddina, none of them fit into the 1 Delitzsch, Paradies, p. 202 seg.; Winckler, Unters., p. 51 seg. > Cf. Hilprecht, Fretbrief Nebukadnezar’s I, and V R. 55-57. ’ On our stone he appears as ‘‘ governor of Bit-Sinmagir ;’’ on that of Za’aleh as “governor of the city of Ishin ;”’ so that he probably had been transferred on the accession of Marduk-nadin-abé, or possibly a little earlier. The pre- vious ‘‘governor of Ishin’? was Shamash-nadin-shumu, son of Atta-ilima (cf. Preibrief Nebukadnezar’s I, col. ii, 17). 4 Gesch., p. 96. ° Gesch., p. 93. © Winckler, Unters., p. 146 seq. 7A cylinder fragment of this king, in possession of Mr. Talcott Williams, of Philadelphia, was transliterated and translated in Z. A. IV, 301-323. Paleographic reasons are decisive in fixing the date of thiscylinder. Mr. Williams has given me his kind permission to publish the cuneiform text in the second part of the present volume. Cf. below, p. 44. 40 OLD BABYLONIAN INSCRIPTIONS remaining characters of the last three names of the dynasty. It follows, therefore, that all the five must have reigned before these. As the kings which have been numbered 4 and 5 are known to have been successors of Mardukn4dinahé, it likewise follows that Nebuchadrezzar I cannot have stood lower than the fourth place in the list. It may be safely asserted, however, that he stood in the first place, and was, therefore, the founder of the Pashe dynasty. To this two objections may be offered: (1) That the traces of the cuneiform characters which follow the number of the years in the Jiist b do not favor the reading of Nabi; (2) that Sayce,' on the evidence furnished by the “ Early Tablet of the Babylonian Chronicle,”? col. IV, 17, claims that place in the list for a king Rammdanu-sharra |or shum|* -cddina. Tn reply to this the fol- lowing is to be said : 1. Scholars have adhered too closely to the view that the mutilated begin- ning of the first line of the List b contains after clu traces of the sign SHU,* the ideogram for the god Marduk. Winckler, in his edition of the list, cuts loose from this assumption, and gives as certain only clu. This variation from the carefully guarded tradition is supported by Bezold’s remark’ that ‘‘at this point the tablet is in a most lamentable condition.” The latter, however, seems to recognize traces of two other wedges immediately following. But the chief probiem is whether beneath the two horizontal wedges of lu, there can be seen a small horizontal wedge so that the sign can be completed to the combination of clu and AG,’ the ideogram for Nabi. From tke fact that all those who have examined the list personally are silent on this point I infer that the tablet at this place is too indistinct to permit any definite conclusion. Then, however, there is nothing in the remaining traces that forbids the reading of Nabi instead of Marduk. 2. From what we know from the scanty cuneiform accounts,’ it is clear that the last years of the Cassite dynasty were a time of war and political disturbance, and that it was the weakness of its last representative which furnished the opportunity for its own overthrow and for the rise of the house of Pashe. No matter what verb may have stood in the effaced passage R. P.’, Vol. VY, p. 112, 1. 16,° the supposition 1R. P.*, Vol. V, p. 112, note 1. 2 R. P2, Vol. V, pp. 106-114. * The reading of the middle character seems to be doubtful. Mr. Pinches would render a great service to Assyriol- ogists by publishing the exact cuneiform text at an early date. 4 Briinnow, J. c., 10834, 6 Z, A. IV, p. 317, note 1. ° Briinnow, J. c., 2786. Cf. Hommel, Gesch., p. 448. 1 Cf. especially R. P.?, Vol. V, pp. 111, 112, 1. 14-22. e ST favor umashshir, ‘he left,’ instead of “‘he renounced”’ or ‘‘abdicated’’ (Pinches). Cf. however, Tiele, 7. c., p. 165. CHIEFLY FROM NIPPUR. 41 of Sayce, that line 17 contains the name of the second king of the Pashe dynasty, seems to me improbable, since the same Hlamite king, Avdin-K hutrutash,' who already had attacked Akkad in the time of Bélshumiddina, is again the assailant in this passage. If Sayce were right, this Hlamite would have made his second incursion into Akkad about twenty years after the first. This in itself is possible, but it is made less proba- ble by the expression ‘‘ RammAnu-shum-iddina returned,” which apparently connects this section closely to that which precedes. Besides it will be noticed that Ramma§- nu-shum-iddina does not bear the title of king, as Bélshumiddina. It seems more probable, therefore, to see in RammAnu-shum-iddina, the unfortunate son (or possi- bly another relative) of Bélshumiddina, who “returned” from the place to which Bélshumiddina or his family had fled, in order to take possession of the throne as his lawful inheritance. This leads me to the discussion of the reasons for regarding Nebuchadrezzar I as the founder of the Pashe dynasty. 1. It needs no proof that at a time when a country is harried by a powerful enemy,” and a descendant of illustrious ancestors puts forward claims to the crown, which are based on historic rights, a usurper who is to found a new dynasty must distinguish himself by eminent courage and ability. Such an able ruler, who, according to our present knowledge, surpassed in preéminence all the other kings of his dynasty, Nebuchadrezzar I is certified to have been. He conducted successfully the wars against Elam, the hereditary enemy of Babylon in the Hast, turned his arms victoriously against the North by “casting down the mighty Lulubean,” and marched, as no other Babylonian king for centuries had ventured, conquering into Syria. 2. It is worthy of notice that both the documents bearing his name are written in connection with his successful conflict with Hlam. His wars with this country, therefore, must have been especially important, perilous and of long duration.’ Since we haye learned from Pinches’ recent publication of the Babylonian Chronicle (col. IV, |. 14-22) that the Elamites took advantage of the weakness of the last Cassite king to devastate Northern and Southern Babylonia, the campaigns of Nebuchadrez- zar I against Hlam become of especial significance. As a usurper he manifestly was able to hold his position only by rendering the Elamites harmless and by defeating them on their own soil, thus “avenging Akkad,” * and restoring quiet and peace to his own country. 1 This and not Khutru ana or Khutrudish (Pinches, l. c., pp. 111-118) is the probable reading. For the value tash of the character in question see Hilprecht in 7. A. VII, pp. 309, 310, 314. .The name means ‘‘subject (servant) of the god Khutrutash”’ (cf. god Martitash). 2K. P.?, Vol. V, pp. 111 seg. * Winckler, Gesch., p. 96. * Hilprecht, Fredbrief, col. I, 13. A. P. §.—VOL. XVIII. F. 42 OLD BABYLONIAN INSCRIPTIONS 3. Nebuchadrezzar I bears titles which differ entirely from those at that time characteristic of the rulers of Babylonia. He calls himself, in the manner of the Egyptians, Shamash mdtishu, “the Sun of his land;” or mushammihw nishishu, “he who makes prosperous his people ;” ndsir kudirétt, mukinu ablé,' “he who protects the boundaries, establishes (measured) tracts of land;” shar kindte sha din mishari idinu, “the king of the right, he who judges a righteous judgment ;” all are titles which probably refer to the fact that just before the reign of Nebuchadrezzar I there was in Babylonia a time of profound misery, when the land did not enjoy sunshine, and when the peaceful possession of well-defined property was impossible, as the violence of the stronger superseded law and order, while, at the same time, the boun- daries of the empire were constantly invaded by powerful enemies; in other words, anarchy as we know it existed in Babylonia at the close of the reign of Bélshum- iddina. The significant title, shdlilu Kashshi, “the conqueror of the Cassites,” acquires doubtless, in this connection, the significance of an allusion to the circum- stance that it was he who had achieved the restoration of the Semitic element through the overthrow of the Cassite dynasty.’ ; 4. The boundary stone IV R.’, 38, which is dated in the time of Merodachbala- dan I, mentions the house (I, 10) and the son (II, 34, 35) of a certain Nazi-Shikhu, while in the “ Freibrief” of Nebuchadrezzar I, a certain Nazi-Shikhu is named as a high dignitary, kalu Akkad. In view of the rare occurrence of this name in Baby- lonian literature ° it is natural to regard the two bearers of the same name as identi- cal. ‘This identification, however, is possible only if Nebuchadrezzar I reigned not long after Merodachbaladan I,‘ 7. ¢., if he, as founder of the Pashe dynasty, came into power some four years after the latter’s death. 1 T formerly transliterated this word aplé (as Peiser still does in Schrader’s K. B. III, Part 1, p. 164). But since i886 I have changed my view and substituted the above. As the word stands parallel to Auduréti, it must have a similar meaning. In spite of nahbalu, II R. 22, 29, b. c., ablé is to be compared with the Hebrew, ‘yan which, in view of the Ethiopic and Arabic hadl has h. Cf. also Delitzsch, Worterbuch, p. 37, no. 30. In view of the title above quoted it does not seem improbable that Nebuchadrezzar I assumed his highly significant name, ‘‘ Nebo, protect the boundary,” only after his usurpation. Another interpretation of the name, ‘‘Nebo, protect (thy) servant,’’ has recently been offered by Jiger (B. A. I, 471, note *). But where is the ‘“‘thy’’? The proper names kudurru and kidinnu, quoted by Jager, (l.c.), are not to be regarded as exclamations but as abbreviations of originally longer names. As the middle part of the name of Nebuchadrezzar is written either kudurru or kudurri (Bezold, Babylonisch-Assyrische Literatur, p. 126), or kudurra (Pl. 82, col. II, 7, of the present volume), it cannot mean ‘‘my boundary,’”’ as I formerly interpreted (Frezbrief, p. viii, note 1), but ‘‘the boundary.’”’ Cf. my remarks in The Sunday School Times, February 20, 1892, p. 115, note 3. 2 Cf. Hommel, Gesch., p. 451. 5 Of. col. VI, 18 of the boundary stone (published by Belser in B. A. II, pp. 171-185), which furnishes us data from the time of the kings Ninib-hudiuré-ugur and Nabt-mukin-aplu. For my transliteration and the formation of the name, cf. above, p. 33 and note 5. ‘For as the son of Nazi-Shikhu who appears as a witness under Merodachbaladan I, was already in possession of the important office of a swkallu, his father must have been advanced in years. CHIEFLY FROM NIPPUR. 43 5. The second king of the Pashe dynasty, according to List b, reigned only six years. And indeed, while the titles and conquests of Nebuchadrezzar I in his “ Frei- brief” imply a comparatively long reign, there are indications that his immediate successor, Bélnadinaplu, ruled but a short time. This does not necessarily follow from the circumstance that the document on Plates 30 and 31 is dated in the fourth year of his reign; but from the fact that Tab-ash&ip-Marduk,' son of Hsagil- zér, already mentioned. under Nebuchadrezzar I as governor of Halw&n, appears again as sukallu in the first year of Marduk-nidin-ahé, 7. ¢., about twenty years later ; for it is very unlikely that the same person occupied a high and responsible position under three successive kings, if both of the former two had reigned a long period. 6. Finally this assumption enables us in the simplest way to dispose of certain chronological difficulties, upon which I cannot enter into details here (cf. e.g. Z A. IONE, [9 208))). The statement of Sennacherib’ furnishes us with a definite datum for the chronol- ogy of the Pashe dynasty. As it seems most natural to connect the carrying off of the images of the gods of Hkallati, with Marduknadinahé’s victory over Assyria, in the tenth year of his reign, we obtain 1107 B. C. as the tenth year of that king’s rule, and 1116 B. C. as the year of his accession to the throne. In accordance with what has been said above, Nebuchadrezzar I reigned 1139-1123 B. C.,° and Bél-nfdin- aplu in 1122-1117 B. C. A word remains to be said as to the length of the period covered by the Pashe dynasty. That the reading of seventy-two years which have been generally assigned to it is impossible, Peiser has shown beyond question by a very simple calculation.® The number of twelve years for the seventh king of this dynasty, assumed by Tiele 1The reading Zabu-ri’éu-Maruduk, ‘‘A beneficent king is Marduk,”’ preferred by Tiele (Gesch., p. 161, note 1), instead of that given above (and first proposed by Oppert and Ménaut in Documents Juridiques), needs no refutation. Tub-aship-Marduk is the only possible one and means ‘‘Good is the exorcism of Marduk.’”’ The Caillow de Michaux upon which Ziéb-ashap-Marduk, apparently not so far advanced in years, likewise appears, belongs to the reign of Nebuchadrezzar I or of Bélnadinaplu (cf. Tiele, J. ¢., p. 161, and Hommel, Gesch., pp. 454, 459). * That Hsagilzér is identical with the Ina-Hsagilzér of the Za’aleh stone (col. II, 12), was shown in my commen- tary on the ‘‘ Freibrief Nebukadnezar’s I,’’ in 1882, which at the time was not printed because of a two years’ illness. At present the proof of their identity is unnecessary. Cf. Hulbar-shurki-iddina, III R. 48, col. I, 29, and Ina-Hulbar- shurki-iddina, V R. 60, col. I, 29. Cf. also Delitzsch, Kossaer, p. 15 (cf. however Gesch., “ Ubersicht’’). Toa dif. ferent effect Jeremias in B. A. I, pp. 270, 280; and Peiser in Schrader’s K. B. III, Part 1, p. 177. 3 Bavian, 48-50. ‘‘Rammin and Sala, the gods of the city of Eka'lati, which Marduknadinahé, king of Akkad, at the time of Tiglath-Pileser, king of Assyria, carried off and brought to Babylon, 1 carried cut of Babylon 418 years later, and brought them back to Ekallati, to their place,’’ 7. ¢., in the year B. C. 689, when Sanherib conquered Babylon. 4 Of. IL, R. 48, col.-I, 5, 27, 28. 5 This calculation confirms strikingly the year 1130 B. C., which I gave as the approximate date of his ‘‘ Frei- brief” in 1883. 6 Z. A. VI, p. 268 seq. 44 OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. (J. ¢., p. 111) and favored by Delitzsch,' finds no support in Winckler’s edition and “besides does not suffice to solve the chronological difficulty. As according to Peiser (1. c.) the passage is much effaced, * and as his proposed reading, 60 + 60 + 12 = 132 years, is the most simple and probable’ solution of the existing difficulty, I accept it and accordingly construct the following table : 1. Nebuchadrezzar I, . . 1139-1123 (seventeen years). 2. Bél-nadin-aplu,. . . . 1122-1117) (six years). 3. Marduk-nadin-ahé, . . 1116-c. 1102 (c. fifteen, at least ten, years). 4, Bape bapien 5. RammAn-aplu-iddina, | ¢- 1101-1053 (forty-nine years). 6-7. Two missing kings, | ; 8. ...........-., . 1052-1031 (twenty-two years). 9. Marduk-bél....., . 1030-1029 (one year and six months). 10. Marduk-zér....., . 1029-1016 (thirteen years). M1. Nabi-shum ....., ~. 1016-1007 (nine years). Total one hundred and thirty-two years and six months. “Anhang’”’ to his Geschichte. 2 Tt is to be regretted that Winckler has not indicated the actual condition of the passage by shading the eflaced portions of the characters. 2 3Cf. also Winckler, Gesch., p. 329, note 17. Another possibility (that 60 + 10+ 10 + 2 —82 stood originally there) is less probable for various reasons. + This name has been transliterated Marduk-shapik-zér-mati (Tiele, Gesch., p. 155 ; Delitzsch, Gesch., “Ubersicht 22) or Marduk-shapik-kul-lat (Winckler, Gesch., p. 98). I regard both transliterations as incorrect, and would substitute that given above for the following reasons: (1) The cylinder fragment published by Dr. Jastrow (cf. above, p. 31, note 7) was unfortunately misunderstood by the latter and misread in various passages. Having examined the frag- ment carefully, I find that the old Babylonian character transliterated ta by Jastrow is distinctly the sign ska in the form so characteristic for the documents of the Pashe dynasty. The name can only be read Marduk-shapik-zi-ri-im. (2) This correct reading is important in connection with the transliteration of the name of Ramman-aplu-iddina’s pre- decessor. It is in itself improbable that two rulers of a Babylonian dynasty of eleven kings bore names almost (if not wholly) identical. The thought forces itself upon our mind that Marduk-shapik-zirim is the same person as the king whose name was heretofore generally read Marduk-shapik-zér-mati. That at least these two names are identical is certain. The last character of the latter name (WAZ, Brinnow, J. c., 7386) was either erroneously read by the Assyri- ologists who copied the so-called ‘‘synchronistic history,’’ or by the Assyrian compiler who used a Babylonian original, instead of the character RIM (Briinnow, 1. c., 8867). For it is well known among Assyriologists that the two charac- ters are nearly identical in the later-middle and the latest periods of Babylonian cuneiform writing. In consideration of this fact, and in view of the phonetic writing z7-ri-im on the cylinder fragment, I unhesitatingly read the name in question either phonetically Marduk-shapik-zir-rim, or ideographically (plus phonetic complement) MWarduk-shaptk- zirim(-rim). The king, Marduk-tabik-zirim, introduced by Dr. Jastrow and accepted by Peiser (Schrader’s K. B. III, Part 1, p. 162 seg.) as an hitherto unknown ruler of the Pashe dynasty thus disappears. As to my other corrections of certain readings offered by Dr. Jastrow in connection with the cylinder in question cf. ‘‘Sprechsaal”’ in one of the next numbers of Z. A. BIBLIOGRAPHY OF THE EXPEDITION. I. Joun P. PETERS. il, . A Brief Statement concerning the Babylonian Expedition sent out under the auspices of the University cw) 2 PP Letter on the Babylonian Expedition: The American Journal of Archeology VII, pp. 472-475. of Pennsylvania : Proceedings of the American Oriental Soctety, April 21-23, 1892, pp. CXLVI-CLUI. Notes on Miirdter-Delitzsch’s Geschichte : Zeitschrift fur Assyriologie VI, pp. 333-339. A Few Ancient Sites, [and IL: The Nation 1889, May 23, p. 423, and May 30, pp. 442, 443. From Niffer to Tello, Land IL: ibidem 1889, July 25, pp. 69, 70, and August 1, pp. 90-92. . Zenobia, Palmyra, and the Arabs : bidem 1890, April 3, pp. 276, 277. . A Misrepresented Ruin : db¢dem 1891, May 7, pp. 375-377. Il. H. VY. HitPRecHT. Rowe ep (Si) . Keilinschriftliche Funde in Niffer: Zeitschrift fiir Assyriologie IV, pp. 164-168. . Aus einem Briefe desselben an C. Bezold : ibidem IV, pp. 282-284. Die jiingsten Ausgrabungen in Babylonien: Kéinische Zeitung 1889, June 30, No. 179. Neue Forschungen in Babylonien: Luthardt’s Hvangelisch Lutherische Kirchenzeitung 1889, June 14, pp. 568, 569. . The Mouth of the Nahr-el-Kelb: The Sunday School Times 1889, Vol. XXXI, No. 11, p. 163. 6. Die Inschriften Nebukadnezar’s im Wadi Brissa: Luthardt’s Zeitschrift fiir kirchliche Wissenschaft und kirchliches Leben 1889 IX, pp. 491-498. Compare also The Sunday School Times 1889, Vol. XXXI, No. 35, pp. 547, 548: The Inscriptions of Nebuchadnezzar in the Wady Brissa. 7. The Shaykh of Zeta: The Sunday School Times 1890, Vol. XXXII, No. 10, pp. 147, 148. 8. Babylon : ibidem 1892, Vol. XXXIV, No. 20, pp. 306-308. . Die Votivinschrift eines nicht erkannten Kassitenkénigs : Zeitschrift fir Assyriologie VII, pp. 305-818. . Kénig Ini-Sin von Ur: zbidem VII, pp. 343-346. . Die Ergiinzung der Namen zweier Kassitenkonige : zbédem, in print. 12. Zur Lapislazuli Frage im Babylonischen : ibidem, in print. IIL Ropert Francis HARPER. 1. ci) Dt D oP oo 10. 11. Babylonian Letter.—The Joseph Shemtob Collection of Babylonian Antiquities, recently purchased for the University of Pennsylvania: Hebraica V, pp. 74-76. . The Kh. Collection of Babylonian Antiquities belonging to the Uniyersity of Pennsylvania : ib¢dem VI, pp. 59, 60. . The Destruction of Antiquities in the East : cb¢dem VI, pp. 225, 226. Three Contract Tablets of Ashuritililani : 2b¢dem VII, p. 79. . A Visit to Zinjirli: The Old and New Testament Student VIL, pp. 1838, 124. . A Visit to Carchemish : ibidem IX, pp. 308, 309. . Down the Euphrates Valley I-III: ibidem X, pp. 55-57; 118, 119 ; 367, 368. . The Expedition of the Babylonian Exploration Fund, A, B. C. : ibidem XIV, pp. 160-165 ; 213-217 ; XY, pp. 12-16; D.: The Biblical World I, pp. 57-62. . Aus einem Briefe desselben an C. Bezold: Zeitschrift fiir Assyriologie IV, pp. 163, 164. Compare also Hebraica VIII, pp. 103, 104: A-bi-e-shu-’ = Ebishum. The Site of Old Baghdad : The Academy 1889, February 28, p. 139. A New Babylonian Contract : 2bédem 1889, April 20, p. 274. IV. THEOPHILUS G. PrncHEs (based upon communications from Dr. Peters and Dr. Harper). 1. An Early Babylonian Inscription from Niffer : Hebroica VI, pp. 55-58. 9 ~ The Discoveries of the American Expedition to Babylonia: The Academy 1891 September 5, p. 199. Compare also his note ‘‘Kadashman :” 7bidem 1891, September 12, p. 221. TABLE OF CONTENTS. Part I, Plates 1-35 and -XY. ABBREVIATIONS. ¢., circa; C. B. M., Catalogue of the Babylonian Museum, University of Pennsylvania; col., column(s) ; d., diameter; Dyn., Dynasty; E., East; fragm., fragment(ary) ; h., height ; Inser., Inscription ; 1., length ; li., line(s) ; m., meter; N., North; Nippur I, II, III, etc., refers to the corresponding numbers on Plate XV ; No., number; Nos., numbers; N. P., Notebook of Dr. Peters made on the ruins of Nippur during the second year’s excavations; Obv., Obverse; orig., original(ly); p., page; Pho., Photograph; Pl., Plate; Rev., Reverse ; S., South ; Sq., Squeeze; T’., Temple of Bél; th., thick(ness) ; W., West; w., width; Z., Ziqqurratu ; Z. A., Zeitschrift fiir Assyriologie. Measurements are given in centimetres. Whenever the object varies in size, the largest measurement is given J. AuTroGRAPH REPRODUCTIONS. Puats. TEx. DatE. DESCRIPTION. 1 1 Sargon I. Door socket in diorite, somewhat smaller than the following. Nippwr III, beneath the rooms of T. on the 8.E. side of Z. Inscr. 18.5 x 10.12, 2 col., 24 li. Sq. On the rough edge, scratched in the rudest way, is the same inser. as Pl.'14, Nos. 23-25 (cf. also Pl. 12, No. 20). Sargon I. Door socket in diorite, 75 x 41.5 x 17.5. Nippur III, same place as No. 1. Inser. 17.8 x 10.35, 2 col., 23 li. C.B.M. 8751. Cf. Pl. I, 1. The variants li. 17 and 21 have been taken from a third door socket in diorite, bearing the same inscr. as No. 2, and found in another trench a short distance from it. Sargon I. Brick stamp of baked clay, brown, with handle, 9.45 x 13.55 x 2. Nippur III, close to the S. E. wall of Z. Inscr. 2 col., 6 li. ‘ C. B. M. 8754. Cf. Pl. IL, 2. 3 4 Naram-Sin. Brick stamp of baked clay, cream colored, handle wanting, 11.75 x 12.08 x 2. Nippur V,in the N.W. extremity. Inscr. 3 li. C.B.M. 8755. Of. Pl. II, 3. 4 5 Al-usharshid. Three fragments of adolomite vase. Orig.d. of the vasec.40. Fragm. 8891 : 11.10 X 7.7 X 3.8. Fragm. 8892 a and b (glued together): 20.5 X 9.8 X 8.8. Nippur IIL, approximately same place as Pl. 1, No.1. Inser. orig. 25.57 X 7.2, 13 li. C. B. M. 8891, 8892 a and b. The text has been restored by the aid of fragm. 8866, 8865, 8843, 8860, 8859, 8858, 8853, 8854 on the scale of fragm. 8892. Cf. Pl. III, 4-12. bo bo (Js) os 48 PLATE. 5 10 TEXT. 6 10 1 15 16 17 18 19 DATE. Al-usharshid. Al-usharshid. Al-usharshid. Al-usharshid. Al-usharshid. Same Period. Same Period. c. 3000 B.C. Ur-Gur. Dungi. Dungi. Ishme-Dagan Ur-Ninib. Bur-Sin I. Bur-Sin IT. OLD BABYLONIAN INSCRIPTIONS DESCRIPTION. Fragm. of a vase in reddish numulite limestone, h. 16.5, d. 18 (of hole 4.4). Nippur IIL, same place as Pl. 4, No. 5. Inser. orig. 11.75 X 7.05,6li. C. B. M. 8888. The text has been restored after No. 5. (Cit, Jel, IVY, 18s : Fragm. of a white marble vase, h. 21, d. 16.4 at the base, 11.2 at the centre. Nippur III, same place as Pl. 4, No. 5. Inser. 4.8.x 5.4, 3li. ©. B. M. 8870. Cf. Pl. V, 14. Fragm. of a white marble vase, orig. h. 6, d. 14.5. Nippur IIT, same place as Pl. 4, No. 5. Inser. (same as Pl. 5, No. 7) 3.2 X 3.8, 3 li. C. B. M. 8839. Fragm. of a white marble vase, orig. h. 13.5, d. 15 (of hole 6.3). Nip- pur III, same place as Pl. 1, No.1. Mark on the bottom, 2.4 x 2.6. Same inser. as Pl. 5, No. 7. N. P. Fragm. of a diorite vase, 7.35 x 2.9 x 0.8, orig. d. 22.2. Nippur III, same place as Pl. 4, No. 5. Inser. 3, orig. 11 li. C. B. M. 8842. White marble tablet, Obv. flat, Rev. rounded, 11.3 X 7.2 X 2.65. Nippur, apparently from the N. W. extremity of V in the neighborhood of Pl. 3, No. 4 (cf. Hilprecht in Z. A. IV, pp. 282-284). Inser. 8 (Obv.) +7 (Rev.)=15 li. C. B. M. 8757. Copied by myself on the ruins of Nippur, April 8, 1889. Fragm. of a large vase in white marble, 10 xX 12.5 x 6.2. Presumably neighborhood of Babylon. Inscr. 2 col., 8 li. C. B. M. 1128. Fragm. of a slab in compact limestone, 12.8 X 7.35 x 5.55. Nippur III, inside of the great S.E. temple wall. Inser. 3 col]., 16 li. C. B. M. 8841. Basalt tablet, Obv. flat, Rev. rounded, lower left corner wanting, 12.25 X 5.58 x 2.2. Northern Babylonia, probably Ursag-Késh. Inser. 8 (Oby.) + 1 (Rev.) =91]i. C. B. M. 841. Agate tablet, bored lengthwise, both sides convex, lower part wanting, 4.4 4.3 x 0.8. Nippur ILI, in a chamber on the edge of the canal outside of the great S.E. wall of T. Oby. Inscr. 8li. C. B. M. 8598. For Rev. see Pl. 21, No. 43. Soapstone tablet, Oby. flat, Rev. rounded, 8.6 x5 X 1.88. Babylonia, probably Mugayyar. Inser. 6 (Obv.) + 2 (Rev.) = 8 li. C. B. M. 842. Fragm. of a slab in diorite, 8.1 x 10.5 x 5.6. Nippur III,S8. of Z. Inser. 3 col.,-3--2+2=71li. C. B. M. 3248. Fragm. of a brick of baked clay, brown, 32 (orig.) X 23 (fragm.) x 8.4 (orig.). Nippur IIL, found out of place in a later structure on the S.E. side of Z. (cf. Pl. 29, No. 82; Pl. 13, No. 22; Pl. 20, No. 38). Inser. (written) 23.3 x 10.65,13 li. C. B. M. 9021. Cf. IV, R. 357, No. 5. Fragm. of a brick of baked clay, brown, 30.5 (fragm.) X 20 (fragm.) x 6.5 (fragm.). Wippur III, found out of place, same place as Pl. 10, No. 18. Inser. (stamped) 22.5 x 10.5, 101i. C. B. M. 8642. Door socket in diorite, an irregular cube, c.19 each side. Nippwr ILI, in a small shrine outside of the great 8S.E. wall of T. Inser. 15.4 x 13.4,2col.,11+6=17li. C. B. M. 8838. CHIEFLY FROM NIPPUR. 49 PratE. DExr: DATE. DESCRIPTION. 13 21 Bur-Sin IT. Door socket in diorite, 33 X 28 X 23. Nippur LIL, same place as Pl. 11, No. 19. Inser. around the hole, 23.5 x 5.35,17 li. Sq. On the bottom at the edge is the same inscr. as Pl. 14, Nos. 28-25 (cf. also Pl. 1, No. 1). 13 22 Bur-Sin II. Brick of baked clay, light brown, very soft, covered with bitumen, 30 < 380 x 6.5. Nippur III, same place as Pl. 11, No. 19. Inser. (written) 5.97 x 5.8, 2 li. Sq. The inscription is generally re- peated three or four times on the same brick (edges and sides). 14 23-25 Gande. Large unhewn blocks of white marble and reddish granite, varying in d. from 25-60. Nippur III, approximately same place as Pl. 1, No. 1. Inser. 6 x 5.3; 7 X 6.2; 6.5 X 7.7; each 3 li. Sq. 15 26 c. 2250 B.C. Cream-colored soapstone tablet, Rev. broken off, 4.85 x 4x 0.8. Pre- sumably neighborhood of Babylon. Inscr. 8 li. C. B. M. 103. 15 27 Hammurabi. Fragm. of an ornamented soapstone stamp in the shape of a vase, h. 18.8, d. 12.2 at the bottom, 8.7 at the centre. Presumably neighborhood of Babylon. Inscr. (on the bottom) 8 li. C. B. M. 11265 @f:, Pl. Txs 20; 15 28 Cassite Dyn. Lapis lazuli disc, d. 1.7. The thickness of this class of inscribed objects found at the same place, if not expressly stated in the following lines, varies from 0.2 to 0.8em. Nippur Ii, same place as Pl. 8, No. 15.. C. B. M. 8685. 15 29 Cassite Dyn. Agate cameo, d. 1.55. Nippur ILL, same place as Pl. 8, No. 15. C. B. M. 8687. 15 30 Cassite Dyn. Lapis lazuli disc, d. 1.6. Nippur ILI, same place as Pl. 8, No. 15. C. B. M. 8721. ; 15 31 Cassite Dyn. Agate cameo, bored lengthwise, 1.7 x 1.9. Nippur III, same place as Pl. 8, No. 15. C. B. M. 8723. 15 32 Cassite Dyn. Lapis lazuli tablet, bored lengthwise, 1.65 x 1.8. Nippur I, apparently out of place, in a gully on the surface. C. B. M. 8720. 16 33 Burna-Buriash. White marble mortar; an uninscribed portion is broken from its side, h. 14.4, d. 12.8. Presumably neighborhood of Babylon. Inscer. $1.5 X 11.25, 271i. ©. B. M.12. Cf. Pl. Ux, 21. 17 383 Burna-Buriash. The same, continued. 18 34 Burna-Buriash. Ivory knob of a sceptre (or conventionalized form of a phallus), top rounded, base flat, round hole in the centre, h. 3.5, d. 5.9 at the top, 6.2 at the bottom. Nippur III, same place as Pl. 8, No. 15. Inser. 5.8 X 2.42, 51i. C. B. M. 8730. Cf. Pl. X, 23. 18 35 Kurigalzu. Tablet in feldspar (mottled dark brown and gray), upper (inscribed) surface convex, lower flat, 3 x 12.2 0.9. Wippur III, same place as Pl. 8, No. 15. Inser. 2li. C. B. M. 8600. ‘ 18 36 Kurigalzu. Irregular block of lapis lazuli, upper part inscribed, 5.1 x 9.25 x 5. Nippur III, same place as Pl. 8, No. 15. Inser. 3.88 x 4.48, 6 li. C. B. M. 8599. Cf. Pl. XT, 25. 19 37 Kurigalzu. Door socket in white marble with red veins here and there, 46.5 x 43.8 22. Nippur III, on the N.E. side of T. near the outer wall. In- ser. on both sides of the hole, 11 li. intended, but only 7 li. inscribed, 14.3 x 14.8. Copied by myself on the ruins of Nippur, April 6, 1889. A. P. S.—VOL. XVIII G. 50 OLD BABYLONIAN INSCRIPTIONS PLATE. TEx. DATE. i DESCRIPTION. 20 38 Kurigalzu. Fragm. of a brick of baked clay, brown, 32 (orig.) x 17 (fragm.) x 7 (orig.). Nippur III, found out cf place in a later structure of the inner wall of Z. (cf. Pl. 29, No. 82; Pl. 10, No. 18). Inser. 13.5 x 6, 9 li, stamped on the edge; the space being too small, a portion of the last character of each line is wanting. C. B. M. 8636. 20 39 Kurigalzu. Fragm. of an axe in imitation of lapis lazuli, 9 x 6.3 x 2.7. Nippur III, same place as Pl. 8, No. 15. Inser. 7li. C. B. M. 9462. Cf. Pl. XI, 26. 21 40 Kurigalzu. Fragm. of an axe in imitation of lapis lazuli, 5 x 6.35 X 1.5. Nippur III, same place as Pl. 8, No. 15. Inser. 4]i, C. B. M. 8661. 21 41 Kurigalzu. Fragm. of a lapis lazuli tablet, 1.7 x 1.7. Nippur LIL, same place as Pl. 8, No. 15. Inser. 3 li. C. B. M. 8662. Originally it formed part of No. 46. 21 42 Kurigalzu. Fragm. of a lapis lazuli tablet, 1.8x 1.2. Nippur III, same place as Pl. 8, No. 15. Inser. 2li. C. B. M. 8663. 21 43 Kurigalzu. Agate tablet. Rev. of Pl. 8, No. 15. Inser. 9 li. 21 44 Kurigalzu. Fragm. of a turquoise tablet. Obv. flat, Rev. rounded; hole bored nearly perpendicular to the lines of the Obv.; 3.4x 3.4 x 0.8. Nippur III, same place as Pl. 8, No. 15. Inscr. 4 li. C. B. M. 8664. 21 45 Kurigalzu. Lapis lazuli tablet, with two holes, 2 x 2.6. Nippur III, same place as Pl. 8, No. 15. Inser. 2li. C. B. M. 8665. 21 46 Kurigalzu. Two fragm. of a lapis lazuli tablet, 3.65 x 7.25. Nippur ILL, same place as Pl. 8, No. 15. Inser, 4li. In cutting the tablet from the original block of lapis lazuli the last characters of each line were lost. C. B. M. 8666. The copy has been made from an electro- type, on which the space between the two fragments was given too small (ef. No. 41). 22 47 Kurigalzu. Nine fragm. of a lapis lazuli tablet, 5.1 x 6 x 0.7. Nippur III, same~ place as Pl. 8, No. 15. Inscr. 6 li. C. B. M. 8667. 22 48 Kurigalzu. Lapis lazuli tablet, hole bored near the top parallel with the lines. 2.8 x 3.45. Nippwr IIL, same place as Pl. 8, No. 15. Inscr. 5 li. C, B. M. 8668. 22 49 Kurigalzu. Lapis lazuli disc, hole kered near the centre parallel with the lines d. 2.5. Nippur IIL, same place as Pl. 8, No. 15. Inscr. 3 li. N. P. 22 50 Kurigalzu. Fragm. of an agate ring, d. 1, w. 0.9. Nippur III, same place as Pl. 8, No. 15. Inser. 5]i. C. B. M. 8669. The ring originally formed the beginning of a votive cylinder (c. 2.6 cm. long), which was afterwards cut in 8 pieces, each thus forming a ring. For the centre part see Pl. 26, No. 74. The last part has not been found. 22 51 Kurigalzu. Agate cameo, 3.2 x 2.4. Nippur III, same place as Pl. 8, No. 15. Inser. 4li. N. P. 22 52 Kurigalzu. Fragm. of an agate cameo, 1.7 1.2. Nippur III, same place as P). 8, No. 15. Inser.2li. C. B. M. 8670. 22 53 Nazi-Maruttash. Fragm. of a lapis lazuli disc, d. 2.97. Nippur III, same place as PI. 8, No. 15. Inser, 6 li, N. P. CHIEFLY FROM NIPPUR. 51 Pirate. Terxr. DATE. DESCRIPTION. 22 54 Nazi-Maruttash. Lapis lazuli disc, d. 2.05. Nippur III, same place as Pl. 8, No. 15, Inser. 5 li. N.P. 22 55 Nazi-Maruttash. Fragm. of an axe in imitation of lapis lazuli, 4.7 x 4.6 x 1.7. Nippur III, same place as Pl. 8, No. 15. Inser.4li. C. B. M. 8671. 23 56 Nazi-Maruttash. Magnesite knob of a sceptre (or conventionalized form of a phallus), top rounded, base flat, round hole in the centre, h. 5.2, d. 6.9. Nippur 111, same place as Pl. 8, No. 15. Inscr. around the top, badly effaced. C. B. M. 8728. Cf. Pl. X, 24. > 23 57 Nazi-Maruttash. Magnesite knob of a sceptre (or conventionalized form of a phallus), top slightly rounded, base flat, hole in the centre (round above, square below), h. 5.2, d. 6.1. Nippur III, same place as PI. 8, No. 15. Inser. around the top, badly effaced. C. B. M. 8727. Cf. Pl. X, 22: 23 58 Nazi-Maruttash. Fragm. of a lapis lazuli disc, d. 4.4. Nippur III, same place as PI. 8, No. 15. Inser. 5 li. (orig. 8). N. P. 23 59 Kadashman-Turgu. Fragm. of a lapis lazuli disc, d. 3.7. Nippur III, same place as PI. 8, No. 15. Inser. 6 li. (orig. 7). N. P. : 23 60 Kadashman-Turgu. Fragm. of a lapis lazuli disc, d. 2.55, -Nippur III, same place as PI. 8, No. 15. Inser. 4 li. Orig. 5). C. B. M. 8722. 23 61 Kadashman-Turgu. Lapis lazuli dise, d. 3.55, th. 0.85. Place unknown, probably Nippur. Inser. 8 li. Original in the Museum of Harvard University, Cambridge, Mass. Cf. Lyon in Proceedings of the American Oriental Society, May, 1889, pp. exxxiv-cxxxvii, and Hilprecht in Z. A. VIL, pp. 305-318. 23 62 Kadashman-Turgu. Lapis lazuli disc, d. 2.7. Nippur III, same place as Pl. 8, No. 15. Inser. 5 li. C. B. M. 8673. 24 63 Kadashman-Turgu. Irregular block of lapis lazuli, 17.5 X 119. Nippur III ina room in the mounds §. of T. Inscr. 16.4 x 9.5, 20 li. Sq. 25 64 Kudur-EN-LIL. Lapis lazuli disc, d. 2.5. Nippur III, same place as Pl. 8, No. 15. Inser. 5li. N. P. 25 - 65 Kadashman-EN-LIL. Fragm. of an agate cameo, d. 3.6. Nippur III, same place as PI. 8, No. 15. Inser. 5 li. C. B. M. 8674. ; 25 66 [Kadashman]-EN- Fragm. of an agate ring, orig. d. 2.7 (of the hole, 0.9), w. 0.96. Nippur LIL. III, same place as Pl. 8, No. 15. C. B. M. 8675. 25 67 [ Kadashman ?]- Fragm. of an agate ring, Rev. of No. 66. Buriash. 25 68 {[Kadashman ?- Irregular block of lapis lazuli, convex on the inscribed surface, BuJriash. 13 X 7.35 X 8. Nippur III, same place as Pl. 8, No. 15. Inser. : 11.5 X 5.9, 3 col., 63 li. (orig. 69 ?). Sq. 25 69 Shagashalti-Shuriash. Magnesite knob of a sceptre (or conventionalized form of a phallus), top rounded, base flat, round hole in the centre, h. c. 5, d. 7. Nippur III, same place as Pl. 8, No. 15. Inser. around the top. Ilo 1B 26 70 Bibeiashu. Magnesite knob of a sceptre (or conventionalized form of a phallus), top rounded, base flat, round hole in the centre, h. 4.6, d. 6.8. Nippur 111, same place as Pl. 8, No. 15. Inser. around the top. C. B. M. 8729. 52 PLATE. 26 26 26 26 26 26 bo ~I 28 TEXT. 71 80 81 OLD BABYLONIAN INSCRIPTIONS DATE. Bibeiashu. [ Bibeia-]shu. Cassite Dyn. Kurigalzu. Cassite Dyn. eMotcsteaete ja-ash. Cassite Dyn. Nazi-Maruttash. [ Bibeia-]shu. c. 1100 B.C. Ramman-shum-usur. Mili-Shikhu. Bél-nadin-aplu. Bél-nadin-aplu. Nabopolassar. DESCRIPTION. Lapis lazuli tablet, 2.35 x 2.16. Nippur Tut, same place as Pl. 8, No. 15. Inser. 5li. C. B. M. 8682. Fragm. of an axe in imitation of lapis lazuli, 11 x 6.95 x 1.25. Nippur III, same place as Pl. 8, No. 15. Inser. 3 li. C. B. M. 8680. Agate cameo, d. c. 1.8. Nippur If1, same place as Pl. 8, No. 15. C. B. M. 8683. Fragm. of an agate ring, d.1, w. 1.1. MNippwr III, same place as PI. 8, No. 15. Inser. 3 li. C. B. M. 8684. The ring originally formed the centre part of a votive cylinder. Cf. Pl. 22, No. 50. Fragm. of an axe in imitation of lapis lazuli, 6x 2.51.5. Nippur IIL, same place as Pl. 8, No. 15. Inscr.6]i. C. B. M. 8681. Fragm. of an axe in imitation of lapis lazuli, 5.26 x 2.1. Nippur III, same place as Pl. 8, No. 15. Inser.4li. N. P. Fragment of a vase in soapstone rock, 8.5 X 8.8 (orig. d. at the bottom 13.2). Nippur V, c. 3 m. below the surface. Inser. 7 li. C. B. M. 8690. Fragm. of an axe in imitation of lapis lazuli, 6.2 x 6.2 1.7. Nippur ILI, same place as Pl. 8, No. 15. Inscr.9 li. C. B. M. 8685. Fragm. of an axe in imitation of lapis lazuli, 2.85 X 2.85 X 1.5. Nippur III, same place as Pl. 8, No. 15. Inser. 4li. C. B. M. 8686. Fragm. of a reddish granite (boundary) stone of phallic shape, I. 15.5. Nippur IIL, ec. 1.5 m. below the surface on the slope of the T. hill N.W. of Z. Inser. 2 col., Sli. Pho. and N. P. Cf. Pl. XII, 32, 33. Fragm. of a baked brick, yellowish, very soft, partly covered with bitumen, 22.5 (fragm.) X 18.4 (fragm.) x 6.9 (orig.). MNippur IIT, found out of place in a later structure of the inner wall of Z. (cf. Pl. 29, No. 82; Pl. 10, No. 18; Pl. 18, No. 22; Pl. 20, No. 38). Inser. written, 15.2 x 8.6,10]i. C. B. M. 8643. Brick of baked clay, brown, partly covered with bitumen, 29.6 x 30.2 X 6.7. Nippur III, inner wall of Z. Every brick of this structure bears the name of Mili-Shikhu with exactly the same inscription (stamped), except a few which belong to Ur-Ninib (Pl. 10, No. 18), Bur-Sin (Pl. 11, No. 19), Kurigalzu (PI. 20, No. 38), Rammanshumusur (Pl. 28, No. 81). The latter four evidently formed a part of the ancient structure, and were utilized by Mili-Shikhu in his restoration of the platform of Z. Inscr. stamped, 14.8x 7, 11 li. C. B. M. 8632. Cf. Pinches ‘‘ An Karly Babylonian Inscription from Niffer”’ in Hebraica V1, pp. 55-58. Black limestone tablet, 16.75 x 12.1 x 5.1. Presumably neighborhood of Babylon. Obv., slightly rounded, 221i. C. B. M. 18. The same, Rev., rounded, 24 li. Cylinder of baked clay, cartridge shaped, hollow, small hole at the top, dark brown with grayish spots; when found, half covered with bitumen; h. 15.2, d. of the base 8.85, d. of the hole 2.2. Babylon. Inscr. 3 col., 45 4+ 65+ 59=169 li. C. B. M. 9090. Cf. Pl. XIII, No. 34. The variants have been taken from a mutilated cylinder (B) in the British Museum, published by Strassmaier in Z. A.V, pp. 129-136. Apparent mistakes in Strassmaier’s edition CHIEFLY FROM NIPPUR. 53 PuatE. TEXT. DaTE. DESCRIPTION. are not quoted as variants (ef. also Strassmaier in Z. A. IV, pp. 106-113, and Winckler in Schrader’s Ketlinschriftliche Bibliothek IIL, Part 2, pp. 2-7). . 33 84 Nabopolassar. The same, continued. 34 85 Nebuchadrezzar II. Fragm. of a baked clay cylinder, barrel shaped, solid, light brown; h. 23.9, d. 8.8 at the top and base, 11.5 at the centre. Babylon. Inscr. 4 col., 23 (orig. c. 48) +32 (orig. c. 56) + 30 (orig. c. 56) + 28 (orig. ce. 48) =113 (orig. c. 208) li. C. B. M. 1785. Cf. Pl. XIV, No. 35. According to information of the Arabs the cylinder was found whole and intentionally broken lengthwise. The other half is supposed to be in existence. 35 85 Nebuchadrezzar II. The same, columns III, IV. II, PHoToGRAPH (HALF-TONE) REPRODUCTIONS. I 1 Sargon I. Door socket in diorite. Nippur. Cf. Pl. 1. IL D Sargon I. Brick stamp of baked clay, Rev. Wippur. Cf. Pl. 3, No. 3. II 3 Nardaim-Sin. Brick stamp of baked clay, Obv. Nippur. Cf. Pl. 3, No. 4. IIL 4-12 Al-usharshid. Fragments of vases from which the text on Pl. 4 has been obtained. Nippur. Nos. 4,5: dolomite; Nos. 6, 8,9, 10: white marble; No. 7: red banded marble of agate structure; Nos. 11, 12: white marble of stalactitic structure. For the restoration of li. 6 fragm. 8860 (white marble) has been consulted. IV 13 Al-usharshid. Fragm. of a vase in reddish numulite limestone. Nippur. Cf. Pl. 5, No. 6. V 14 Al-usharshid. Fragm. of a white marble vase with gray and reddish veins here and there. Nippur. Cf. Pl. 5, No. 7. VI 15 Not later than 2400 B.C. Fragm. of a white marble slab, 26.65 x 15.8 x 7.9. Abu Habba. Orig. inal in Constantinople. Photograph taken from a cast. Inscr. on both sides and left edge, 391 li. Oby., 9 col., (20 + 254 24 + 224 22 +26 + 19 + 23 +4—=) 185 li. VIL 16 Not later than 2400 B.C. The same, Rev., 9 col., (19-4 19 + 23 + 25+ 28 + 24 + 25 + 92 +13 =) 198 li. VIII 17 Not later than 2400 B.C. The same, left edge, 1 col., 18 li. VIIIL 18, 19 c. 2400 B.C. Tablets of baked clay, reddish brown with black spots. These tab- lets have a peculiar shape; they are rounded at both ends and on the left side, but angular and flat on the right side, as if cut off from a larger tablet. Yokha. No. 18: 10.3 x 4.3, th. 1.6 on the left, 2.2 on the rightside. C. B. M. 9042. No. 19: 10.62 x 4.5, th. 1.7 on the left, 2.55 on the right side. C. B. M. 9041. Ix 20 Hammurabi. Fragm. of an ornamented stamp in the shape of a vase, made of soap- stone (composed of a green micaceous and very soft mineral, prob- ably tale). Presumably neighborhood of Babylon. Cf. Pl. 15, No. 27. XII 29-31 OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. DATE. Burna-Buriash. . Burna-Buriash. Nazi-Maruttash. Kurigalzu. Kurigalzu. c. 1350 B.C. c. 13850 B.C. c. 1150 B.C. ce. 1100 B.C. Nabopolassar. Nebuchadrezzar II. 1889 A.D. DESCRIPTION. Fragm. of a white marble mortar. Presumably neighborhood of Babylon. Cf. Plates 16, 17. Knob of a sceptre (or conventionalized form of a phallus) in ivory. Side view. Nippw. Cf. Pl. 18, No. 34. Knobs of sceptres (cf. Pl]. X, 23) in magnesite. Cf. Pl. 23, Nos. 57, 56. Inscribed block of lapis lazuli, tablet in process of cutting. Nippur. Cf. Pl. 18, No. 36. Fragm. of a votive battle axe in imitation of lapis lazuli (blue glass). Nippur. Cf. Pl. 20, No. 39. Fragm. of a votive battle axe in imitation of lapis lazuli, 8.32 x 5.65 x 5.1. Nippur III, same place as Pl. 8, No. 15. C. B. M. 8800. Fragm. of a votive battle axe in lapis lazuli, 6.4 5.71.5. The inscription has been erased in order to use the material. Nippur III, same place as Pl. 8, No. 15. C. B. M. 8597. Three small fragments of an inscribed bas-relief in a basaltic stone, h. c.5. Nippur III, on the S8.E. side of the Bur-Sin shrine (cf. Pl. 11, No. 19). Fragm. of a reddish granite (boundary) stone of phallic shape. Nip- pur. Two views of the same stone. Cf. Pl. 27, No. 80. ; Cylinder of baked clay, cartridge-shaped, hollow, small hole at the top. Babylon. Cf. Plates 32, 33. Cylinder of baked clay, barrel-shaped, solid. Babylon. 34, 35. Plan of the first year’s excavations at Nippur (February 5 to April 16). Top views. Nippur. Cf. Plates ARTICLE II. THE MAMMALIA OF THE DEEP RIVER BEDS. BY W. B. SCOTT, COLLEGE OF NEW JERSEY, PRINCETON. Read before the American Philosophical Society, October 6, 1893. The ninth Princeton geological expedition to the Tertiary formations of the Far West selected as its field of operations the valley of Smith river, or Deep creek, as it is variously called, in central Montana. The party, which was under the direction of the writer and Prof. W. F. Magie, consisted of the following students: Messrs. Butler, Benet, Coulter, Hosford, Jefferson and Stevenson, and spent a part of the summer of 1891 in exploring the very limited outcrops of lacustrine beds in the region mentioned. We had the good fortune to secure the services of Prof. O. C. Mortson, of Great Falls, Mont., as a guide, and to his minute knowledge of the country and zealous labors the success of the undertaking is in large measure due. Many gentlemen in Great Falls, White Sulphur Springs and Livingston took great interest in the work of the expedition and rendered every assistance in their power. To enumerate all of those to whom we are under obligations for many kind- nesses would be impossible, but special thanks are due to the Hon. Paris Gibson and Mr. W. W. Connor, of Great Falls, but for whose most kind and prompt assistance at a critical period the trip would necessarily have been abandoned. GroLogicaL Musrum, Princeton, N. J., September 20, 1893. The literature of the Deep River or “ Zicholeptus” beds is rather limited, as the region has been comparatively little explored. The formation was first discovered by Grinnell and Dana, in 1875, and their brief account may be quoted almost in full. “During the explorations carried on last summer under the direction of Col. William Ludlow, Corps of Engineers, a series of Tertiary deposits were identified by the writers near Camp Baker, Montana. These deposits indicate the existence in A. P. S.—VOL. XVIII. H. 56 THE MAMMALIA OF THE DEEP RIVER BEDS. this region of a Miocene lake basin, which was succeeded by another lake basin in Pliocene time. “Camp Baker [the spot marked Logan on the map herewith given] is situated on Deep creek, a stream which flows into the Missouri river above Sun river. It lies about fifty miles due east of Helena. It is surrounded on all sides by mountains, of which the Big Belt range, lying immediately to the south or southwest, is the highest and most conspicuous. .... “The Tertiary beds found here consist, for the most part, of homogeneous cream- colored clays, so hard as to be with difficulty cut with a knife. he beds are hori- zontal and rest unconformably upon the upturned red and yellow slates below. The clays of which they are formed resemble closely those found in the Miocene [z. e., White River] beds at Scott's Bluffs, near the North Platte river, in Wyoming. The deposits at Camp Baker have been extensively denuded and nowhere reach any great thickness. At a point about three miles southeast of the Post, some bluffs were noticed where the Miocene beds attained a thickness of about two hundred feet, and these were capped by fifty feet of the Pliocene clays, both beds containing character- istic fossils. In the underlying Miocene beds were found a species of Rhinoceros, several species of Oreodon Leidy and Hporeodon Marsh, a canine tooth apparently of Elothertum Pomel and remains of Turtles. In the Pliocene beds the principal fossils were a species apparently of Merychyus Leidy, remains of an equine smaller than the modern horse, and Pliocene Turtles. These fossils have not yet been carefully studied, and for this reason their relations to the remains found in the other lake basins of similar age cannot be stated. “We saw the first exposures of these beds a few miles west of the Sulphur SJOMUMIGISS 6 5 56 This point is about six miles southeast of Camp Baker. From here, the bed was traced continuously along Deep creck for a distance of fifteen miles, extending quite up to the mountains, on the eastern side at least. Beds of the same character, containing similar fossils, were found on White-Tailed Deer creek, a branch of Deep creek, about seven miles to the north of Camp Baker, as well as on Camas creek, to the southwest of the Post. Traces of this deposit, containing what, appear to be remains of R/inoceros, were also found two miles or more south of Moss Agate Springs and at a considerable clevation above the creek bed. With more time than we had at command, they could no doubt have been traced much farther, although in many places the beds have been washed out or have been covered by the later local drift. “These Tertiary beds were ail laid down after the elevation of the mountains and the igneous eruptions. They are, as has been said, perfectly horizontal, and are THE MAMMALIA OF THE DEEP RIVER BEDS. 57 often seen covering over ridges of trachyte. The line of separation between the Miocene and Pliocene beds is in some places well marked. It consists of about six feet of hard sands, interstratified with layers of very small water-worn pebbles soldered together into a hard mass, but very easily picked out with a knife. Imame- diately above the strata of pebbles the Pliocene fossils were found. In several places fragments of trachyte were noticed in the Pliocene beds” (No. 18, pp. 126-128). The next account of this formation was given by Cope, who had sent his assist- ant, Isaacs, to collect in the valley and described a number of new forms from it. His collection embraced specimens from the upper beds only, those called Pliocene by Grinnell and Dana, and these he referred to the Loup Fork. In 1879 (No. 3), Cope divided the Loup Fork into two horizons, which he called the Zicholeptus and Procamelus beds respectively, the former being the beds of the Deep River region. Subsequently (Nos. 6, 8, 10), Cope raised the Ticholeptus beds to a rank coordinate with that of the John Day or the Loup Fork, and gives the following list of species as occurring in the Montana area: Mastodon proavus, Protohippus sejunctus, Mery- cocherus montanus, Merychyus zygomaticus, M. pariogonus, Oyclopidius simus, C. emylinus, Pithecistes brevifacies, P. decedens, P. heterodon, Procamelus vel Protolabis sp. Blastomeryx borealis (No. 8, p. 369). In his latest paper on the subject, this writer defines the formation as follows: “Ticnotuprus. Mammalia. Presence of Anchitherium, Proboscidea and Camelidz and the Oreodont genera Merycocheerus, Merychyus, Cyclopidius and Pithecistes. Absence of ? Hlotheriide, ? Poebrotherii- de, ? Nimravidee and Cosoryx. This horizon requires further exploration, as but twenty species have been thus far determined from it. But it is evidently interme- diate in age between the John Day and Loup Fork epochs, with greater affinities to the latter. It differs from the latter in the presence of Anchitherium, numerous genera and species of Oreodontide, and in the absence of Cosoryx. The formation is known from three regions: first, from western Nebraska; second, from the valley of Deep river, Montana; and third, from Cottonwood creek, Oregon. Its thickness has not yet been stated” (No. 6, pp. 456, 457). Tt should be noted that in these lists the name Anchitherium is used for the John Day equines, to which, in this paper, I have applied Marsh’s name, Miohippus, for reasons which will appear later. This point is of importance. Z Tn 1891, I published a brief note upon the subject of this horizon (No. 31). At that time the fossils collected were still in the matrix, and only the hasty examina- tions in the field were available for the purposes of comparison. Consequently, a number of errors crept into the work, so as to greatly vitiate its conclusions, which will not be further referred to here. In a second note (American Naturalist, 1893, 58 THE MAMMALIA OF THE DEEP RIVER BEDS. p- 660) I gave preliminary definitions of the new genera and species contained in the collections made by the Princeton party. Now that this collection has been worked over and can be compared with Prof. Cope’s material from the same locality, some definite statements may be made with regard to the geological and paleontological relations of the Deep River beds. So far as the stratigraphy is concerned, there is little to add to the account of Grinnell and Dana, except in one particular. The statements of these authors seem to imply that the two sets of beds are conformable throughout, but there is strong evidence which goes to show that this is not the case. In the first place, there is a marked litholog- ical contrast between the two series, the lower being very hard and the upper, for the most part, incoherent sands, though nodules of harder material have, in many cases, formed around the bones. The general character of the lower beds is very much like. that of the older Miocene, the White River or John Day, while the upper are more like the Loup Fork. Though both sets of strata are generally horizontal, with local exceptions, the upper beds appear to rest upon an eroded surface of the underlying strata. Thus, at one point, the older beds, as exposed in a line of buttes—appar- ently, at least—rise higher than an exposure of the newer strata across the ravine from the first exposure. In the absence of instruments, this point could not be determined quite certainly, but it is very probable. ‘Towards the north and east the upper beds appear to extend beyond the lower and to produce an uncomformity by overlap. Finally, the fossil contents of the two series of strata are very strikingly different, not a single species of mammal and not more than two genera are common to the two, and those genera range from the John Day into the upper Loup Fork. Such radical and sudden changes are hardly to be explained on the hypothesis of migration, and point to a considerable hiatus between the times of deposition of the two sets of strata. The following species of mammals were found in the lower beds: Cynodesmus thocides Scott, Steneofiber montanus Scott, Cenopus sp., Miohippus annectens ? Marsh, M. anceps? Marsh, M. (Anchitherium) equiceps ? Cope, Mesoreodon chelonyx Scott, M. intermedius Scott, Poebrotherium sp., Hypertragulus calcaratus Cope. This list appears to be a scanty one, but this is explained by the fact that the exposures which yielded well-preserved fossils are very limited in extent, a few acres at most, and when we compare them with the vast regions over which collections from the other Tertiary formations have been gathered, the disproportion will not seem so striking. Indeed, I know of very few spots of equal extent which have yielded so large a number of individuals and species. The facies of this fauna is undeniably that of the John Day Miocene. All of the genera but two, and seyeral of the species, occur THE MAMMALIA OF THE DEEP RIVER BEDS. 59 in the typical Oregon localities, and while, owing to the very small area of the Mon- tana beds, we cannot lay much stress upon the absence of certain characteristic Jobn Day forms, yet the presence of such relatively modernized genera as Cynodesmus and Mesoreodon indicates that these beds should be referred to the summit of the John Day formation. This is of interest as being the first identification of this hori- zon east of the Rocky mountains. The upper beds, which Grinnell and Dana called Pliocene, present a very difter- ent assemblage of species. Cope’s collection, so far as I can judge, was gathered entirely from these beds and contains nothing from the lower horizon. His collection and that made by the Princeton party are, as would naturally be expected, not quite coéxtensive, each containing some forms which the other does not. Combining the two, we obtain the following list: Canis? anceps Scott, Chalicotherium? sp., Aphe- lops sp., Miohippus sp., Anchitherium equinum Scott, Desmatippus crenidens Scott, Protohippus sejunctus Cope, Protohippus (Merychippus) insignis Leidy, Merychyus (Ticholeptus) zygomaticus Cope, M. pariogonus Cope, Merycocherus montanus Cope, Cyclopidius simus Cope, C. emydinus Cope, C. incisivus Scott, Prthecrstes brevifacies Cope, P. decedens Cope, P. heterodon Cope, Protolabis sp., Procamelus sp., Blastomeryx borealis Cope, B. antelopinus Scott, Mastodon proavus Cope. In addition to this list should be mentioned a considerable number of equine animals, which cannot be well identified, as the specimens are scattered vertebre, limb and foot bones, not accompanied by teeth, but which, from the variations in size and details of construction, point to several species not enumerated above. The resemblance of this fauna to that of the Loup Fork has been obvious from the first, for it was doubtless the latter formation to which Grinnell and Dana referred under the name “Pliocene.” For the same reason of very limited exposures, as in the case of the lower beds, when compared with the John Day of Oregon, we cannot insist very strongly upon the absence of typical Loup Fork genera from the upper series of Deep River strata. Of much greater significance is the occurrence in the latter of five genera and fourteen species of mammals which have not been found in the vastly more extensive and carefully examined Loup Fork deposits. This fact, having regard to the character of the species involved, points to the conclusion, already drawn by Cope, that these beds are older than the typical Loup Fork hori- zon, but their faunal connection with that horizon is so close that there seems little ground for considering the Deep River as an “epoch” of coordinate rank with the three other Miocene epochs. The relation between the Deep River and Loup Fork beds is more intimate than that between the Wind River and the Bridger proper of the Hocene. My own preference is, therefore, to refer both series to the Loup Fork, 60 THE MAMMALIA OF THE DEEP RIVER BEDS. as Cope originally did, and then subdivide that formation into two horizons. The names which Cope first proposed for these subdivisions, the Zicholeptus and Proca- melus beds respectively, are inapplicable, for the former name is a synonym of Mery- chyus, a genus which occurs in both horizons, and, as now appears, Procamelus prob- ably does also. I cannot agree with Cope in regarding the strata of western Nebraska and Cot- tonwood creek, Oregon, as referable to the same horizon as those of the Deep river valley, in Montana. In the case of the former, the determination rests chiefly upon the presence there of Leptauchenia, which Hayden found associated with Oreodon, Ischyromys, Hyracodon and other characteristic White River forms (see Leidy, No. 23, pp- 20, 21). Cope has questioned the correctness of this statement as to Leptau- chenia, but it has been abundantly confirmed, that genus being an undoubted White River form. Hayden’s reference of Merycocherus and Protomeryx to this same hori- zon is almost certainly erroneous and has not been confirmed by subsequent observ- ers. The reference of the beds developed along Cottonwood creek and the upper John Day river, in Oregon, to the Deep River horizon, is determined by the occur- rence in them of a so-called Anchithertum and of a species identified as Blastomeryx borealis. It should be noted, however, that the term Anchitheriwm is used in the sense of Miohippus, the species from Montana which I have called A. equinum is avery different animal and belongs to the group of A. aurelianense, of Europe, which it equals in size. Mvohippus is found in the typical Loup Fork, as well as in the lower series (see Osborn, No. 28, p. 89, under the title ?Anchitherium parvulum). No great weight, therefore, can be attached to the occurrence of the genus in the Cot- tonwood Creek beds. The presence of Blastomeryx borealis would, of itself, be insufficient for the correlation of the two localities, but the identification of the species is not at all certain. Besides certain minor differences in the teeth, the limb bones from the Oregon beds indicate the existence there of two species, both of which are much heavier than the Montana forms and more like others from the Loup Fork of Kansas. Cope, himself, was struck by the faunal differences of the two localities. He says: ‘‘The only species common to the two lists is the Blastomeryx borealis, a fact which indicates some important differences in the two horizons, either epochal or faunal” (No. 8, p. 369). Present evidence appears, therefore, to point to the conclusion that the upper series of strata developed in the valley of Deep river form a well-marked horizon at the base of the Loup Fork, and that they are not exactly paralleled by any deposits as yet known cisewhere; and, further, that the lower series of the Montana strata should be referred to the summit of the John Day, where they form a less distinctly TIE MAMMALIA OF THE DEEP RIVER BEDS. 61 marked horizon. Together, the two series tend to bridge over the gap between the John Day, on the one hand, and the typical Loup Fork, though they by no means completely close the hiatus. It is a more difficult task to correlate these beds with their European equivalents. The Loup Fork horizon was referred by Leidy and Hayden to the Pliocene, a view which is still maintained by some authorities, but, as Cope has shown, the determination rests upon the supposed occurrence in these beds of forms having a very modern facies, and which were very probably derived from newer overlying strata, since, in typical Loup Fork exposures not covered by these newer beds, the modernized forms have not been found. The recent discovery of the Blanco beds, of Texas, with their true Pliocene fauna (see Cope, No. 11), lends additional force to Cope’s contention that the Loup Fork should be referred to the upper Miocene. Branco has objected to this correlation, as follows: “‘ Wine scharfe Parallelisirung wird hier durch die verschiedene Zusammensetzung der beiderseitigen Faunen erschwert. Auf der einen Seite fehlen der Loup Fork Gruppe echt miocaene Formen wie Anthracother‘um und Anchither‘um und es treten dafiir Geschlechter von jugendlicherem Aussehen wie Protohippus und Hippidium auf. Andererseits aber repriisentiren nicht nur die amerikanischen Oreodontide ungefiihr ein mit dem europiiischen Cwenothertum ibereinstimmendes Entwicklungs- stadium, sondern beiden Faunen sind auch direct Steneofiber, Amphicyon, Tetraloph- odon, Hipparion und Procervulus gemeinsam. Man wird also mit Cope diese Paral- lelisirung der Loup-Fork-Gruppe mit dem Miocaen Huropa’s im Allgemeinen gelten lassen miissen, wenn gleich man nicht iibersehen darf, dass dieselbe durch Formen wie Protohippus und Hippidium, welche dem Pferde der Jetztzeit bereits recht naheste- hen, sowie durch das Vorkommen von Dicotyles, Hystrix und Mustela einen entschie- den jugendlicheren Charakter erhiilt als die miocaene Fauna Europa’s” (No. 2, p. 149). These objections rest, for the most part, upon the incorrect identifications of European and American genera, which were current at the time Branco’s paper was written. As will be seen in the sequel, Anchither’um is present in the lower Loup Fork and not in the White River and John Day, the equines of which formations have been erroneously referred to that genus. The absence of Anthracothervum from the Loup Fork is of no weight, since the genus is quite unknown in America. The occurrence of Hystria, Dicotyles and Mustela in the Loup Fork beds is extremely doubtful, the identifications being made on very imperfect specimens. The reference of Hippidium to this horizon is also very doubtful and has not been confirmed. If, as is almost certainly the case, the equine series is of American origin, there is noth- ing surprising in the fact that the series should be, on this continent, one stage in 62 THE MAMMALIA OF THE DEEP RIVER BEDS. advance of its contemporaries in Europe, just as the American ruminants in several horizons lag behind their contemporaries of the Old World. Further, to regard the Loup Fork as Pliocene involves the assertion that Procamelus in America was con- temporary with Camelus in Asia, which, seeing that the camel series is of American origin, is most improbable. Of the European faunas, that of Sansan and Simorre offers the best analogy with that of the Deep River beds. In both continents this horizon is marked by the first appearance of the mastodons, and, since the Proboscidea would seem to have originated in neither Europe nor America, but to have reached both regions by migration, this fact is significant. -Anchitherium, in the restricted sense in which I have used that word, is likewise common to both, and, as this genus has a very restricted range in time, it is a most important fossil. Blastomerya is exceedingly like the Sansan species of Palwomeryx, though more primitive in some respects. No stress can be laid upon the supposed Amphicyon and Procervulus of the Loup Fork, as these names are incorrectly given to the American forms. If Dall’s contention, that North and South America were not united until the close of the Miocene (No. 12, p. 21), be confirmed, the Loup Fork will necessarily be referred to the Pliocene, as is indicated by the occurrence in those beds of the glyptodont genus, Caryoderma Cope, a South American type, though the genus itself has not yet been obtained in that continent. But the evidence for the date of the elevation of Central America and the Isthmus of Panama is by no means con- clusive. Gabb’s statement is to the effect that “The communication between the Atlantic and Pacific in the region of Costa Rica was interrupted in the Pliocene or subsequent to the deposition of the mass of the Miocene strata” (italics mine) (quoted in Dall and Harris, No. 12, p. 188, from Gabb’s MS. report). This is quite compati- ble with the view that the connection of the two continents was made before the end of the Miocene. If we may provisionally regard the Deep River beds (upper series) as equivalent to those of Sansan and Steinheim, the John Day would consequently be about equivalent to the lower Miocene of St. Gérand le Puy, though probably somewhat older, and the White River to the beds of Rouzon or the “ Marnes lacus- tres.” The term Oligocene has not been found necessary in this country, the line between the Uinta and White River beds being a clear and convenient demareation between the Eocene and Miocene. Nevertheless, much confusion and incorrect rea- soning haye resulted from calling the White River simply Miocene. The presence of such genera as [Zycnodon, Hemipsalodon, Mesonyx, Hlotherium, Hyopotamus, ete., sufficiently proves these beds to be more ancient than the true Miocene of Hurope, THE MAMMALIA OF THE DEEP RIVER BEDS. : 63 and it would be much to the advantage of clearness and consistency if the White River were called Oligocene, a view for which Cope has long contended. In the descriptive portion of this paper the fossils will be treated according to their systematic position, as the interest attaching to them is especially morphologi- eal. The lack of smaller animals in the collections is very noticeable. The lower beds have yielded but one rodent and the upper none at all; only two carnivores, both dogs, have been found. None of the Insectivora or Chiroptera have been detected. The fauna, as so far known, consists, therefore, almost entirely of medium and large-sized ungulates, for which the conditions of fossilization are, no doubt, chiefly responsible. CARNIVORA. Canide. CYNODESMUS Scott. : (Pl. I, Figs. 1-5.) Amer, Naturalist, 1893, p. 660. : Canine animals having the dentition of the microdont forms of Canis, but with the skull structure of the ancient genera. Cerebral hemispheres small, not overlap- ping the olfactory lobes or cerebellum, and with fewer and simpler convolutions than any of the recent Canidw. Postglenoid foramen concealed or absent. CYNODESMUS THOOIDES Scott. (loc. cit.) Dentition microdont ; deuterocone of upper sectorial relatively well developed ; face short, cranium elongate; small frontal sinuses present; mandible non-lobate with stout angular hook and broad, recurved coronoid; size medium. The technical distinction of this genus from Cynodictis is by no means easy, and yet it becomes very clear on an examination of the two; while the latter very probably represents a side branch, leading away from the direct canine phylum, the former may, with equal probability, be regarded as being either in the direct line of canine descent or but little removed from it. In order to make clear the character of this interesting form, it will be most convenient to compare it carefully with some typical modern species, for which pur- pose the coyote, Oanis latrans, will be taken as a standard. I. Dentition. A. Upper Jaw. The incisors are very small and form a nearly straight row, the external pair projecting but little behind the others. The first and AS PB. S:=—ViOlL. X Viti. I. 64 THE MAMMALIA OF THE DEEP RIVER BEDS. second are of nearly equal size, while the third is somewhat larger, though the differ- ence is much less marked than in C. latrans or even than in the John Day species, Temnocyon corypheus ; the teeth are crowded together, not spaced apart as in the latter species. ‘The diastema between the external incisor and the canine is rather short. The latter is as well developed, relatively, as in the coyote, but of a some- what different form, being more oval and less compressed in section; the depression on the inner face and its anterior bounding ridge of enamel are also less marked. Except for their relative shortness (antero-posteriorly ) and height, the premolars closely resemble those of the modern species. P.1 is a very small and simple tooth, implanted by a single fang; the crown is of compressed conical shape, without pos- terior cingulum. P.2 is much larger, though small as compared with the same tooth in C. latrans ; it is of elongated, compressed, conical form, but has no posterior basal cusp (tritocone) such as occurs in the coyote. P.3 is a still larger tooth of similar construction, except that a small tritocone and posterior cingulum have been added, which, however, are less conspicuous than in C latrans. The sectorial (p 4) differs but little in any respect from that of the existing microdont species of Canis; the protocone exceeds the tritocone less in antero-posterior extent than in the coyote, and the deuterocone is much more distinctly developed than in that species, so that the transverse diameter of the crown is greater, not only in proportion to the antero-pos- terior diameter (length) and to the size of the whole skull, but actually as well. In some recent species of Canis, however, the deuterocone is quite as well developed. The premolars are quite closely crowded together and set obliquely to the line of the alveolus, so as to slightly overlap one another; posteriorly, the two lines of premolars diverge quite rapidly, while the molars converge, so that the angulation between the two series is very marked. The first molar is wider proportionately to its fore and aft length than in C. latrvans, and the external cusps, para- and metacones are lower, more conical, and less angulate and pyramidal in shape than in the coyote. The cingulum is very strongly developed at the antero-external angle of the crown, soas almost to deserve the name of a parastyle, while it becomes very faintly marked upon the metacone. ‘The inner elements of the crown, the protocone and crescentic cingulum, are not nearly so prominent as in the coyote; the anterior conule is slightly better developed, and the posterior distinctly less so, than in that animal. M.2 is much reduced; the external cingulum is faintly marked, except on the paracone, and the internal one not nearly so strongly developed as in Canis latrans. B. Lower Jaw. The incisors are very small and set closely together, and, as is usual in the dogs, the second pair are crowded back out of line with the others; in THE MAMMALIA OF THE DEEP RIVER BEDS. 65 dimensions, they increase regularly from the first to the third. The median pair are too much worn to show whether they possessed bifid crowns, but this is clearly the case in the second and lateral pair; the posterior groove which indicates this struc- ture is more median in position than in the modern form. The canines are shaped very much as in the latter, but are shorter and diverge less towards their apices. The first premolar resembles the corresponding upper tooth, having a very small and simple crown supported ona single root. ‘The succeeding premolars increase regularly in size up to the fourth; essentially they are all alike—compressed, trench- ant, acute and quite high cones; on p.3, and more distinctly on p.4, a posterior basal cusp and cingulum appear, but they are less developed than in © latrans. The first, or sectorial molar, is characteristically cynoid, but retains some primi- tive features. Compared with the inferior sectorial of the coyote, the following dif- ferences are apparent: (1) The protoconid is relatively higher, less compressed, and more conical in shape, shorter in the fore and aft dimension, and its anterior border is much more steeply inclined and nearly vertical; (2) the paraconid is lower and less extended antero-posteriorly ; (3) the talon is lower, and, while it is as broad as the anterior portion of the crown (trigonid), and therefore entirely different from that of Temnocyon, yet its basin-like character is less emphasized than that of Canis, owing to the smaller size and less elevation of the entoconid; the metaconid corresponds in size and position to that of Canis. The differences enumerated are slight and yet not without importance; for whenever the sectorial of Cynodesmus departs in struc- ture from that of Canis, it is in the direction of Daphenus and the creodonts. M2 differs in no tangible respect from that of the coyote. M.3 is not so much reduced as in that species and has a more elongate oval crown, which is supported on two fangs, while, in the recent representatives of the family, the fang is very generally single. II. Tue Sxuut. The skull preserves many of the primitive characters which occur in the ancient genera, such as Temnocyon and Daphenus. This is particularly marked in the long, narrow cranium, with postorbital constriction placed far back of the orbits, and the short face, which is due partly to the microdont dentition and the anterior position of the orbits, they being farther forward than in Canis. The basi- cranial axis, as measured by Huxley’s method (No. 19, p. 239), is strikingly long, actually exceeding that of the considerably larger skull of C. latrans. This elonga- tion of the cranium does not, however, imply a correspondingly long cerebral fossa, as may be seen from the position of the postorbital constriction, which marks the anterior boundary of the hemispheres, and which, in this genus, is much farther remoyed from the orbits than in the recent members of the family, in which it follows 66 THE MAMMALIA OF THE DEEP RIVER BEDS. close behind the orbits. The cerebral fossa is not only narrow, therefore, but it is short, extending only slightly above the cerebellar fossa, and the lengthening of the basicranial axis more particularly affects the floor of the latter. In the fossil which is under description, the roof of the brain-case and the occiput, together with the condyles, have been weathered away, and therefore the conformation of the sagittal and occipital crests cannot be determined with certainty; but, from the character of the frontal ridges and the shape of the cranial cast, which is well preserved, there can be little doubt that these crests were very much as in Zemnocyon corypheus, to the skull of which species that of Cynodesmus bears a very close resemblance. The upper contour of the skull is nearly straight and the descent at the forehead very slight and gradual, in which respect we find a great similarity in shape to the fox’s skull. The basioccipital (so much of it as is preserved) is narrower than in Canis, broader and more flattened than in Zemnocyon ; in the latter, this bone is anteriorly much narrowed by the extremely large bulls, and posteriorly displays a median longitudinal convexity, with a deep fossa on each side of it. In the species before us, the paroccipital process is very different from that of Canzs ; in the latter it is “long and prominent, and its anterior surface is applied closely to the back part of the bulla, but to a less extent than in the cats, as the process is more compressed. The mastoid is distinct but slightly developed” (Flower, No. 14, p. 24). In Cyno- desmus, as in Temnocyon, the paroccipital process is much longer, more compressed, and more curved downward and backward; its free portion is much more widely sep- arated from the bulla, with which the process is connected by a narrow bridge of bone, which expands anteriorly so that the contact surface between the two is about as in the existing genus. The mastoid is somewhat more exposed on the sur- face of the cranium than in Canis and is more lateral in position, the paroccipital processes occupying the inferior angles of the occiput. This displaces the mastoid processes anteriorly, so that, as in Temnocyon, they are on the sides of the skull and overlapped by the squamosal; they are somewhat more developed than in Canzs. The tympanics are inflated into large auditory bulle, which equal in actual size, and therefore proportionally exceed, those of Canis latrans, though they are much less prominent than in Temnocyon corypheus. So far as can be judged from the speci- men, the bulla appears to be divided by a septum, in very much the same manner as in Canis, into two widely communicating chambers, of which the postero-internal is much the larger. The meatus auditorius is an irregularly oval opening, which does a? not form a tube; the anterior lip is, however, extended outward more than in Canis and, separated only by a narrow slit from the postglenoid process of the squamosal, articulates with it at its extremity. The shape and development of the bulla produce THE MAMMALIA OF THE DEEP RIVER BEDS. 67 some differences in the disposition of the foramina in its neighborhood; thus, the foramen lacerum posterius extends less around the hinder end of the bulla and is confined to its postero-internal angle, running almost parallel to the basicranial axis. Still more important is the fact that the anterior lip of the auditory meatus, extend- ing along the postglenoid process, overlaps and conceals the glenoid foramen. In- deed, I cannot altogether satisfy myself that the foramen is present at all; but there is a long, narrow and curved slit between the lip and the process, which probably contains the entrance to the foramen. In Zemnocyon coryphwus the foramen occupies the position of the slit just mentioned, but is much more conspicuous, resembling the same structure in the raccoon. The zygomatic arches are relatively longer and more massive than in Cans ; they arch outward as far, but much less strongly upward, and thus, when seen from the side, pursue a straighter course. The root of the zygomatic process of the squamosal is continued backward as a broad shelf over the mastoid process, as is also the case in Lemnocyon corypheus though not in Canis. The glenoid cavity is more extended transversely than in the latter, and is more concave, the hinder margin being elevated into a ridge, which rises gradually into the postglenoid process, which is longer and more curved anteriorly than in the coyote. The jugal is very long and extends backward to the outer angle of the glenoid cavity; the masscter ridge is more prominent and rugose, and the masseter surface wider than in the recent animal ; the postorbital angle, which is but slightly developed in the latter, does not appear at all. The anterior end of the jugal is bifurcate and the inferior branch descends lower upon the molar alveolus than in the coyote. The lachrymal has about the same extent upon the face as in that species, but possesses a spine in the form of an obtuse ridge; the foramen is single and placed entirely within the orbit. The specimen does not permit us to determine the share taken by the frontals in forming the roof of the cranium, but they possess considerable extension upon the face. The supraciliary ridges are well marked and rugose and converge rapidly to form the sagittal crest; clearly, no lyrate “sagittal area” could have been present. The forehead is not so flattened as in Zemnocyon, but slightly arched from side to side, and the postorbital processes are hardly more developed than in that form and consequently much less so than in Canis latrans. The nasal processes are very long and nearly reach the premaxillaries, though in this respect there is some asymmetry in the specimen, the process on the left side being appreciably longer than that on the right. Small frontal sinuses are present. The nasals are relatively long, and are broader and more convex from side to side than in the coyote; the anterior border is not emarginate, but obliquely truncate and censiderably longer than in the recent 63 THE MAMMALIA OF THE DEEP RIVER BEDS. animal. The anterior nares are higher, more oval in shape and more inclined back- ward than in Temnocyon corypheus, less so than in Canis latrans. In the former they are small, nearly circular in shape and vertical in position. The horizontal por- tion of the premaxille is shorter, less massive and rounded than in the coyote, in correlation with the smaller incisors, and at the symphysis the two are less closely applied. The ascending portion is also quite differently shaped; it is much longer, broader and more steeply inclined, and its superior and anterior borders pass into each other almost imperceptibly, while in the coyote the two meet at an angle not very much greater than a right angle. The palatal portion differs but little from that seen in the latter species, but the incisive foramina are somewhat more anterior in position and encroach less upon the maxillaries. The maxillary, in its extension upon the face, is short, but relatively deep verti- cally, and this height rapidly increases backward, so that the premaxillary suture is steeply inclined. The canine alveoli cause more marked prominences upon the face than in Canis latrans, and the muzzle is more constricted behind them. The infira- orbital foramen is nearer to the orbit than in that species, but occupies the same position with reference to the teeth, opening above the interval between p.3 and p.4. The palatal processes are somewhat narrower than in the coyote, and the suture between them is marked by a low rugose ridge. The palatines have a less extent, both in length and breadth, than in the modern form, their anterior borders, which in the latter reach to the interval between p.8 and p.4, hardly extending beyond the mid- dle of the sectorial. On the other hand, the front margin of the posterior nares is quite behind the molar alveoli, while in the coyote it is opposite the front of m.2; the palatal notches are also much less deeply marked than in the latter. The pos- terior nares are long and narrow and somewhat constricted in the middle of their course; the pterygoids have larger hamular processes than in the coyote and the pterygoid fossz are better marked. The mandible differs in important respects from that of Canis latrans. The hori- zontal ramus is shorter, but deeper and thicker; the chin rises more steeply, which produces less procumbeney in the incisors; the lower border is more sinuously curved, descending more abruptly from beneath the coronoid, and the angular hook longer and stouter. The ascending ramus has a greater antero-posterior extent, and the coronoid is broader, more inclined backward, and with more curved posterior margin; its anterior border is wider and more distinctly defined and displays a groove for the attachment of the buccinator and maxillo-labial muscles, which would seem to indicate that these muscles were better developed than in the existing form. This broad anterior surface is reflected over upon the upper border, where it forms a THE MAMMALIA OF THE DEEP RIVER BEDS. 69 very distinctly marked, flattened and obliquely inclined surface for the insertion of the temporal muscle; its upper margin, however, is a thin edge, not thickened and rugose as in the coyote; its lower margin forms the upper boundary of the masse- teric fossa and is continuous with the prominent ridge which bounds that depression anteriorly. The nearest approximation to this character of the coronoid which I have been able to find among the recent Canide occurs in C. cinereo-argentatus. The masseteric fossa is large and profound, indicating a powerful muscle, which is further confirmed by the character of the surface on the jugal for the origin of the masseter. The condyle is somewhat flattened upon its postero-superior aspect ; it is much more extended transversely than in Canis latrans, and this extension is most marked in the portion external to the coronoid. The cranial foramina, with the exception of the foramen lacerum posterius and the glenoid foramen, which have already been noticed, depart in no respect from those of Canis. The mandible has a large mental foramen beneath p.2 and a smaller one under p.3, which are closer together than in the coyote; the dental foramen occupies the same position as in that species. Ill. The Brain. The cranial cast displays characters very different from those of the recent Oanide, both in its general proportions and in the details of the cere- bral convolutions. The hemispheres are narrow in proportion to their length and taper gradually forward; their contour is rather more alopecoid than thooid, accord- ing to Huxley’s distinctions. ‘In the Fox the contour of the brain, viewed from above, is that of a pear with the narrow end forwards, laterally the contour is undu- lated, presenting one slight incurvation in the region of the sylvian sulcus and another in that of the supraorbital [¢. é., presylvian] sulcus, while a little angulation marks the junction of the olfactory lobes with the cerebral hemispheres. In Canzs azarw the cerebral hemispheres immediately behind the supraorbital fissure widen out abruptly and the lateral contour, instead of being slightly incurved at this point, presents a sharp rectangular inflection. The frontal lobe anterior to the supraorbital sulcus is much longer in C. azare than in C. vulpes and the brain is considerably wider behind in the latter” (No. 19, pp. 245-247). In Cynodesmus, the posterior widening of the alopecoid brain does not occur, but the anterior portion is more like what occurs in those animals than in the thooids, though simpler than in either. The hemispheres slightly overlap the lateral lobes of the cerebellum, but are notched in the middle, so as to leave the vermis free. Owing to the relatively well-developed tempero-sphenoidal lobes, the cerebrum has considerable vertical depth in this region, but anteriorly it is very shallow as well as narrow. Apparently, the hem- ispheres leave the olfactory lobes quite exposed. Except for its greater width pos- 70 THE MAMMALIA OF THE DEEP RIVER BEDS. teriorly, the brain of the fennee (Canis zerda) has a very similar outline to that of the fossil. We can best examine the sulci of the hemispheres after quoting Krueg’s description of the fissures which are characteristic of the recent Canide: “ Fissura anterior und postica sind immer vorhanden, fast immer verbunden. Fissura coronalis, ansata, lateralis, medilateralis meistens verbunden, letztere sowie die ectolateralis immer vorhanden. Die Fissura splenialis ist hiiufig mit der rhinalis posterior, nur ausnahmsweise nicht mit der cruciata verbunden. Die Fissura prorea, precruciata, postcruciata und confinis fehlen hiufig und sind auch bei den grosseren Species nie stark entwickelt” (No. 21, p. 614). In Cynodesmus the cerebral convolutions are much simpler than in any existing species of Canzde, even the smallest. Besides the sylvian fissure, the dorsal aspect of the hemispheres displays but two slightly curved sulci, one of which, the superior, is clearly the lateral sulcus; its anterior portion may, perhaps, represent the ansate and coronal fissures, but if so, all three are in the same straight line. In the recent species the three are usually connected, but with the difference that the ansate and coronal sulci are curved downwards and forwards, out of the line taken by the Jateral. The second fissure in the specimen is the supra- sylvian, which is remarkably short and little curved, and is not continued into the posterior suprasylvian, which appears to be absent. The crucial fissure is not indi- cated on the cast, but no great stress can be laid upon this fact, for this sulcus is sometimes not shown in the intracranial casts of recent species, the brains of which actually possess it. If present, however, in Cynodesmus, it must have been extremely short, as is shown by the straight course of the lateral sulcus and its nearness to the dividing fissure between the two hemispheres. Among several brain casts of Miocene carnivores, I have seen none which displays the crucial sulcus, though we can scarcely believe that this fissure, which is now so characteristic of the recent families of the order, had not then been developed. One very striking difference between the cerebral sulci of Cynodesmus and those of the existing dogs, is the absence in the former of the posterior prolongation and downward curvature of the fissures. The medilateral is lacking, and a minute, isolated depression is all that can represent the ectolateral. Nor do I find any trace of the presylvian (supraorbital) sulcus, or of the “fissura anterior und postica,” which, in the recent species, are always present and nearly always connected to form a strongly curved sulcus between the sylvian and the suprasylvian. It is of interest to note that this brain, in its simplicity of convolution, is much more like that of foetal dogs than of any adult recent species. Among existing carnivora, we find such simple sulci approximated only in some of the smaller viverrines and mustelines. THE MAMMALIA OF THE DEEP RIVER BEDS. 71 The cerebellum does not differ in any important respect from that of recent dogs, except that it is less extensively covered by the hemispheres. The vermis is promi- nent and well convoluted, and is principally extended upon the dorsal side. The lateral lobes are somewhat injured by weathering, so that the degree of their convo- lution is not apparent. In the subjoined table, comparative measurements are given of the skulls of Cynodesmus, Temnocyon corypheus and Canis latrans. Under each species, the first column gives the actual dimensions in millimeters, while in the second column are the proportionate measures, by Huxley’s method, the length of the basicranial axis in each case being taken as 100. These figures are calculated only to the nearest integer, the fractions representing amounts which are far within the limits of indi- vidual variation. CYNODESMUS. C. LATRANS. T. CORYPHAUS. NL AGS = 100, M. Axts=100. M. Axis=100. Length of basicranial axis............... 066 100 061 100 .066 100 otalelen= thio fasicullteeeeee eet eet 147 223 161 264 153 234 en'othpoferaceaaria-sectetcee eo cie eerie 048 73 .073 120 .056 85 ene thwotazyCOmamer- repo aeee eee 074 112 072 119 074 112 Width across zygomas...............---- .106 161 .098 161 cass ae Length upper molar-premolar series...... .052 79 .069 113 055 83 Length of upper sectorial................ -015 23 020 33 017 26 Width of upper sectorial................ O11 ily -009 15 -O11 15 Length of first upper molar............... 011 17 -013 21 -010 15 Width of first upper molar.............. .016 24 O17 28 016 24 Length of second upper molar........... 007 . 11 008 13 004 6 Width of second upper molar............ .010 15 012 20 -009 i4 AWildthwofapalatera tis 4crrrerterterl-reliettarelers -043 65 052 85 -050 WG Length of auditory bulla co0dn0b000 0006 024 By 023 38 027 39 Mente theo femandibleneerciaieis ase) ler siele-et-l- 113 171 -130 212 Length of first lower molar.............. 017 26 -021 34 Length of second lower molar........... .008 12 010 16 Length of third lower molar............ 005 8 -005 8 Tt should be added that, in Cynodesmus, the length given for the basicranial axis is approximate only, the margins of the foramen magnum being broken away, but the error cannot be sufficient to detract from the substantial accuracy of the results. This beautiful specimen was found in the lower beds of Deep River valley by Prof. O. C. Mortson. Tur SystemATic PostriIon of CyNODESMUS. Of the phylogeny of the Canidw, which has so long remained obscure and puzz- ling, Schlosser says: ‘Die Abstammung des Hundes ist noch immer mehr oder weniger in Dunkel gehiillt. Es sind zwar eine grosse Menge fossiler Fleischfresser A. P. S.—VOL. XVII. J. 72 THE MAMMALIA OF THE DEEP RIVER BEDS. bekannt, die jedenfalls in naherer oder entfernter Beziehung zu dieser Gruppe stehen, allein da von denselben das Skelet entweder noch nicht gefunden ist oder doch von jenem der Hunde sehr bedeutend abweicht, so bleiben wir noch immer iiber die eigentlichen Ahnen des Hundegeschlechtes fast ganz und gar im Ungewissen” (No. 29, p. 247). The gradual recovery of the dogs of the American Miocene formations is bringing us nearer to a satisfactory solution of this difficult problem. As Schlosser has pointed out, the numerous cynoids of the Huropean Oligocene, with the possible exception of some species of Cynodictis, can be of little phylogenetic significance, and in the lower Miocene of Europe the dogs disappear completely ; they are repre- sented in the upper Miocene and Pliocene by only a few remains, and do not attain great importance till the Pleistocene (No. 30, p. 488). Throughout the American Miocenes, however, from the White River to the Loup Fork, they piay a very impor- tant part, and are not only abundant in individuals, but very varied in type, no less than nine genera of Miocene dogs, most of them containing several species, having been described from the different American horizons. This fact of itself would indi- cate the greater probability of an American rather than a European origin of the family. In the Loup Fork beds, aside from the aberrant /lurodon, several species of cynoids occur which are indistinguishable from Canis, and, so far as the remains at present known are concerned, must be referred to that genus, though complete material will probably require their separation from it. One of these species, C. brachypus Cope, is very probably of phylogenetic importance and is significantly like Cynodesmus. This is a microdont species, which retains many primitive char- acters, such as the small sectorials, short face, long cranium, elevated sagittal crest and weak feet. The lower Loup Fork strata of the Deep River valley (Cyclopidius beds) contain a possible species of Canis, the O.? anceps, which will be hereafter described. So far as this form is known, it is intermediate between the C. brachypus and Cynodesmus. The latter genus is found in the lower Deep River beds, which we have already referred to the summit of the John Day horizon, and its connection with C. brachypus is a fairly close one, as is apparent not only in the dentition but in the characters of the skull as well; as, for example, in the characteristic shape and con- nections of the paroccipital processes, length of the zygomatic arches, size and shape of the coronoid process of the mandible, ete. We may also fairly assume that “ the elevated sagittal crest and the small feet” are shared by the older genus. The White River type, Daphenus, is separated from Cynodesmus by a wider interval, the typical John Day horizon, in Oregon, not having as yet yielded any form which can be placed in the series, unless we are to find the missing link in the species THE MAMMALIA OF THE DEEP RIVER BEDS. 73 referred provisionally to Zemnocyon by Cope, under the name of 7. josephi ; but until the mandibular dentition of that species is discovered, its place in the eynoid series must remain indeterminate. In spite, however, of the considerable gap between Cynodesmus and Daphenus, their relationship is indicated by nearly every detail of known structure in the two genera. The skull characters are closely similar in both ; e. g, the long, narrow cranium, with postorbital constriction placed far back of the orbits, and the short, rapidly tapering face. Cynodesmus shows advance over Daph- ceenus in the following particulars: (1) The auditory bulle are enlarged, fully ossified and the posterior chamber indistinguishably fused with the anterior, while in the White River genus the posterior chamber remains cartilaginous, or, at all events, is separate from the anterior. In all the skulls which I have had the opportunity of examining, the posterior chamber is lost, exposing the periotic from below, and the anterior chamber is very small. (2) The cranium is somewhat more rounded and capacious, and it, together with the zygomatic arches and ascending ramus of the mandible, has become somewhat shortened. (3) The sectorials are rather more mod- ernized and efficient shearing blades, the cusps being more compressed and extended and less conical in form; in p.4 the deuterocone is reduced. (4) The third upper molar has disappeared. (5) The first upper molar has become smaller and the outer cusps moved nearer to the edge of the crown. Temnocyon represents a slightly modified side branch, in which the inferior sec- torial has developed a trenchant talon, through the reduction or suppression of the entoconid. In all other respects, the true canine character of Temnocyon and its close resemblance in skull structure to Cynodesmus are very striking. Whether T. corypheus and T. altigents are properly placed in the same genus, or whether, as Schlosser suggests, they belong to widely separated phyla, need not be discussed here, nor can we determine at present whether Zemnocyon and Cynodesmus have any common ancestor nearer than Daphenus. Icticyon alone, among recent dogs, shares with Temnocyon the character of the trenchant talon on the inferior sectorial. As this character is a rare one, both in fossil and recent cynoids, we may, perhaps, expect that the existing South American genus will prove to be derived from the John Day type. If so, many intermediate forms remain to be discovered. Daphenus, in the structure of its skull, dentition and limbs, approximates closely to the creodonts. This approximation is seen in the character of the secto- rials, which are very like those of the Macide, in the primitive form of the cranium, in the low humeral trochlea with its epicondylar foramen, in the third trochanter of the femur, the creodont-like caleaneum and the relatively weak plantigrade feet. The eynoids of the Uinta formation are, unfortunately, not sufficiently well known for 74 THE MAMMALIA OF THE DEEP RIVER BEDS. generic or even ordinal reference, as it is uncertain whether they are carnivores or creodonts, but it is altogether likely that they will prove to be intermediate between Daphenus and the Miacide of the Bridger. The great difficulty in the way of making out a satisfactory phylogeny of the Canide is the position to be assigned to the problematical genus Otocyon. If, as so many authorities maintain, it be inadmissible to assume that in this form the number of teeth has been increased at a comparatively recent period and within the limits of the family, then we shall be driven to admit a very remarkable degree of parallelism, or rather of convergence. Hither the series of fossil forms which lead by slight and natural gradations from the Macide to Canis have nothing to do with existing spe- cies, but merely form a parallel series, leading to no permanent result, while the real — ancestors of the family are entirely unknown; or, on the other hand, Otocyon must represent the te1mination of a line leading upwards from some creodont family, as yet undiscovered, which line has paralleled the dogs in every detail of structure except the dentition. For my own part, I am by no means convinced of the impos- sibility of the addition of new teeth to the molar series. That modification in the mammalian lines is very generally by way of reduction in the number of teeth, is true, but does not prove that the reverse process may not exceptionally take place, whether by reversion or otherwise. The great simplicity of the teeth in Otocyon can hardly be reconciled with its advance in all other respects, except on the hypothesis of a retrogression or reversion in dental structure. At all events, such an assump- tion would seem to involve less of improbability than either horn of the dilemma to which its rejection confines us. Stress has been laid upon the lyrate sagittal area of Olocyon and its occurrence in the young of other species of the family as showing that it is a primitive charac- ter. But an examination of a series of fossils in almost any mammalian phylum shows that the high and thin sagittal crest is the primitive character, and its replace- ment by a flattened area the secondary modification. The reason for this is plain; in the ancient forms, the jaws and canine teeth are powerful and the brain is small, hence the cranium does not offer sufficient surface for the attachment of the tem- poral muscles, and the sagittal crest must be developed, just as in the analogous case of the sternal keel in birds. Now, the disproportionately large size of the brain in the young animal gives a large surface for muscular attachment at a period when the weak jaws and small milk teeth require little muscular power, and hence the devel- opment of the crest is retarded. In no embryonic structure are there so many “cenogenetic” features as the skull, just on account of the great and premature enlargement of the neryous axis and the higher sense organs, and hence embryologi- THE MAMMALIA OF THE DEEP RIVER BEDS. 75 cal data must be applied with great caution in discussing questions of skull morphol- ogy. For the same reason, the sagittal crest is relatively less developed in very small forms, as is exemplified in the small species of Cynodictis from the White River and John Day, and in the least of these, C. (Galecynus) lemur, a lyrate area is formed, the most ancient known cynoid in which it occurs. A sagittal area also occurs in the White River insectivores, Leptictis and Jctops, but it is worthy of note that it is not marked in the older species of the latter, from the base of the Bridger. There are equally good reasons for regarding the lobate mandible which is found -in many of the recent Canide@ as a secondary modification. It not only is absent in all Miocene members of the family and all known creodonts, but, so far as I can ascertain, it occurs only in those recent species in which the mandibular condyle is much elevated above the level of the molars, and this is by no means a primitive character. The following table will serve to display the relationships of the various Ameri- can genera of the cynoid stem, so far as the available material renders this possible: Loup Fork, . 6 . ?Canis, Allurodon. John Day, : ; Cynodesmus, Temnocyon, Cynodictis. White River, : » Daphenus, Cynodictis. Uinta, ¥ : é ? Bridger, F 6 eieaces: CANIS. 2? CANIS ANCEPS Scott. Amer. Naturalist, 1893, p. 660. A small fragment of mandible, containing the last premolar, first and second molars, is provisionally referred to this genus. It agrees well with Cope’s descrip- tion of Canis brachypus from the Loup Fork (No. 4, p. 389), except for its inferior size and relatively more slender mandibular ramus. The inferior sectorial is nearly as long as in that species, in the proportion of 17 to 19, but the depth of the jaw beneath that tooth is much less, as 21 to 30. Possibly the species should be referred to Cynodesmus, but several minute details point rather to Canis. 76 THE MAMMALIA OF THE DEEP RIVER BEDS. The sectorial is very like that of Cynodesmus ; it is marked by a very short (antero-posteriorly) paraconid and high conical protoconid, which has a very steeply inclined anterior edge; the talon, however, is somewhat more modernized by the increase in size of the entoconid, which bears about the same proportion to the hypo- conid as in Canis latrans, and, as in that species, a minute cusp is present between the base of the metaconid and the entoconid, which does not occur in Cynodesmus. The second molar is like that of C. latrans, except for the larger size of the ento- conid, which in the modern species is reduced to a mere ridge. The third molar is missing, but its alveolus shows it to have been larger than in the coyote, and the fragment of mandible displays a deeper and thicker horizontal ramus than in that species. Measurements. engthwofsfirstelowersmolars(antero posterion) see eee rene rere eee ner Eeee ere ree eee eee ereree 017 ene vhwofssccondulowersmolarseen eee ECee eee E ee eer Ee cOE eee ee reerEEer ree ae ers ». .008 ILegN OF Thiel loner wml ATCO, coc cn0o0ccup ons obosdunavanononoanooonsonsocesegosene -006 Depthyorsmandibleabeneathpm'plep eee Eeee eee eee ee aee err e eee ere EE CREE EEEC Eee eer rer nrran 021 The type specimen of this species was found inthe Cyclupidius beds of the Deep River by Mr. I. Benet. RODENTIA. Castorida. STEN EOFIBER. This genus may be distinguished from the nearly allied Castor not only by the simpler pattern and less markedly prismatic character of the molar teeth, but also by the absence of codssification between the fibula and tibia. In the American spe- cies the humerus always has an epicondylar foramen. STENEOFIBER MONTANUS Scott. Amer. Naturalist, 1898, p. 660. This species is most like the 8. (Custor) peninsulatus Cope, from the typical John Day horizon, but differs from it in some details of molar construction. In the upper molars, except m. 8, there are but two fossettes, both of which are anterior to the external enamel inflection; the latter also is nearer to the posterior border of the crown than in the other species of the genus. In the lower molars there is, as in the other species, “a deep external enamel inflection and three transverse lakes on the inner side ;” but different from any of the other members of the group, the external THE MAMMALIA OF THE DEEP RIVER BEDS. V7 inflection is hardly at all oblique in its course, but runs nearly parallel to the fore and aft axis of the tooth. In both of the lower teeth represented in the specimen, the antero-posterior diameter of the crown exceeds the transverse, which is very unusual in this genus. The incisors are narrow, with anterior faces which are less convex than in Castor and are covered with a thick layer of orange-colored enamel. Two caudal vertebree indicate that this species had a longer and more slender tail than the beaver; the anterior portion was provided with chevron bones. The humerus has a rather slender, trihedral shaft and prominent deltoid ridge, which terminates in a massive overhanging hook; this hook is proportionately even better developed than in Castor. The supinator ridge is also conspicuous and con- tinues high up upon the posterior aspect of the shaft. The trochlea is low and nar- row, more so than in the beaver, but otherwise shaped as in that animal, and the anconeal fossa is very shallow, not so deep, in fact, as the supratrochlear. The inter- nal epicondyle is very prominent, massive and rugose, and is perforated by a large foramen. ' The femoral trochanters are well developed, but the third is placed more proxi- mally than in 8. peninsulatus or in the beaver. The calcaneum has a short, depressed, irregular and club-shaped tuber; the sustentaculum is notably smaller than in the modern species, and the external projection near the distal end much more prominent; the cuboidal surface is of triangular outline and slightly concave. Of the metatarsals only the third is preserved in the specimen, but this is sufficient to show that in this species of Steneofiber, at least, the foot had very different propor- tions from what we find in the existing genus. This metatarsal is relatively very much more slender and shorter than in Castor and of quite different shape, as the shaft is of nearly uniform size throughout, not being contracted in the middle nor expanded distally ; it is also more depressed and flattened, and the head for the first phalanx less enlarged. The proximal end has an oblique surface for the ectocunei- form, which is abruptly constricted behind and continued as a narrow posterior tongue. This specimen suggests very strongly that when Steneofiber becomes com- pletely known it will prove to be much better distinguished from Castor than the skull and dentition have led us to suppose. Measurements. M. Length Of first Upper MOAT. 3.26.0 .55. cee eke ee tere ee ee ee te esteem ence etc sane sess .004 VAG OE tala, Wyse wie, ooasnnneeocvsdcoo eco onoabon gD OOO OR OOOO MOOE OU OD OED DOOD OR HOnODDDO .005 Length of second upper molar.......... 0. ese eee teen eee nett cette tec erence es 004 Width of second wpper MOAT... 1.6... nec eeee eee eee eee cence ete eect eee tee sees tees 005 Length of third upper molar..... 2... 02... c eee eet e eee eee eet e nee ete tenet eee see case 003 78 THE MAMMALIA OF THE DEEP RIVER BEDS. Width: ofp thirdjuppers molar vi. .j5- ooncccnsecopcosoooncconedc0cconnsoon0nGe GOOF 014 Wippenmmolanssenieswslent (hese ener ener reset iertiesetitsttt ert te ets. 055 .053 Wippersirstomol are Lemig thy. je.) tafelajer=telaeicl eke eistel ited te oleh teste letel-\-t- cir ten-taiatod 015 012 Whoyoer ithas5 avo, Wills occpoaacoonncobendogs Doo docu no COsBROSAHODaESS C000 015 Ujpoer eaconal MOP, NONI, coc09os06s505 6050005000 n00as 0c9000RancadaKe .018 .016 Won saConGl ime, WACIWN, ocondcadnocaconacadc 00a nods a0DDs00000000000 C000 -U21 Wjooer tiniiel molar; Wea, ocasccsoscoass coven scoccnbossece cooBaneaGeoRS .025 024 Whose Wail more, TAN oocaccacooH cee acaonoccDoveOKDD OOO UDYOBOOUDaGOO O00 -025 Lower molar-premolar series, length....-......-.-.-+2+---+.--.sseeeseeee 107 -109 Wowerspremolariseriess lenetheree tear eee eee eee reteset .048 046 Owershits tepremolaryle nie Uhre ele aeee Cece er attri irr 010 008 LOK ier ite: jOREmMOEN, WihNe ocsocoocenaocba pcos occu 004nb00e00G0000000 009 006 A. P. S.—VOL. XVIII. Q. 128 THE MAMMALIA OF THE DEEP RIVER BEDS. MALE. FEMALE. M, M. owerssecondepremolarslengthsneeseeceretaseeceece ease reece .013 013 Howersthindspremolaniwden|s Ubi sper ter acre ti wee rere 013 -014 owernourth=spremolar.slen oer pete eee eee ee Eee eee 015 014 Lower molar series; lengths sniecacscts eros oe ete eT eee ears 059 064 Lower? firsthmolar length pees yasei aoe oe ere i oleae ee ease eee .014 014 Howerssecondemolarglen tthe eee erence eee aera 017 018 lowersthirdymolarolen alec era reek een eee eee .028 032 ~The Milk Dentition is like that of the older genera in that 4.3, which in Mery- chyus is molariform and composed of four crescents, is like neither molar nor pre- molar. The anterior half of the crown is a compressed protocone with trenchant edges, like the corresponding cusp in the premolars, but thicker transversely, and with a fossette upon its anterior face. Except for its greater thickness, this portion of the crown is like the whole of 4.2 or the corresponding premolar. The posterior part of the crown is composed of a pair of transversely placed crescents, the trito- and tetartocones, and resembles half of a molar. As I have elsewhere pointed out, with reference to the more ancient members of the family, this tooth plainly shows that in the upper milk molars the homologies of the cusps, as determined by their position, are the same as in the premolars, but the order in which these cusps appear is altogether peculiar, being as follows: proto-, trito-, tetarto- and deuterocones (No. 29, p.441). In the lower series, 4.2 and 4.3 are like their successors in the permanent dentition, while 4.4 is of the usual artiodactyl pattern, consisting of three pairs of crescents; of these, the anterior pair is formed by the paraconid and an element internal to it, to which, as it occurs only among the artiodactyls, I have not thought it worth while to give a special name. I cannot determine whether p.1 has a prede-" cessor in the milk series, as it has in Oreodon, though there is some reason to think that this is the case. If so, the change takes place at an early period, before any of the other milk molars are shed. The milk canines and incisors differ from the per- manent ones merely in size. In short, the temporary dentition of Mesoreodon departs from that of Merychyus more widely than does the permanent one, though in this connection it should be remembered that the temporary teeth are not known in the earlier species of this genus, /. zygomaticus and M. pariogonus, and as these species have a permanent dentition which in one or the other respect recalls that of the older genera, it may well happen that the temporary dentition of these species will also prove to be inter- mediate between that of the typical Merychyus species from the upper Loup Fork and that of the more ancient forms of the family. The Skull (Pl. IV, Fig. 32; Pl. V, Fig. 35). The structure of the skull is so THE MAMMALIA OF THE DEEP RIVER BEDS. 129 like that in Hporeodon that no detailed description of it will be necessary, and it will therefore suffice to mention the points of difference between the two genera. The general proportions of the skull, length of face and cranium, size and position of the orbits, ete., are very similar in both; it is only when we come to compare the details that differences become apparent. In Mesoreodon the anterior aspect of the premax- illaries is slightly broader and more flattened, and the two bones are more closely applied together and the symphyseal portion is more thickened. These changes are very slight, but they are not unimportant, since they are in the direction of the curious ankylosed premaxillaries of Merychyus. The maxillary sinuses are enlarged, which gives to the face a slightly swollen appearance. As in the older genera of the family, there appears to be a sexual difference in the lachrymal depression, the depth of which varies with the size of the canines, indicating that it was better marked in the males, but it is never so deep as in the males of Hporeodon. The frontal sinuses are more inflated than in the John Day genus, which gives to the forehead somewhat the same vaulted appearance as in Merychyus, but to a less degree; the nasal pro- cesses of the frontals are unusually long. The zygomatic process of the squamosal is intermediate in character between that of Hporeodon and that of Merychyus zygomaticus ; it is very widely expanded at the base, both transversely and antero-posteriorly, and its outer border is quite strongly raised, thickened and rugose, more so than in the former, less so than in the latter. The postglenoid process resembles that in the John Day form in being low, broad and very massive. The tympanic bulle vary in size, being in some specimens much more prominent and inflated than in others. So far as the material in hand goes, it appears to indicate that the bulle were more largely inflated in the male than in the female, but a much larger series of skulls will be required to definitely deter- mine whether this is really a sexual character or not. The occiput is peculiar, and in its upper portion very similar to that of Hporeodon, the angles being extended into a pair of large wing-like processes as in that genus and in some species of Orcodon as well. These prominent processes are separated by a deep concavity; beneath this the surface is transversely very convex, prominent in the median line, and with deep grooves or narrow fossz at the sutures between the squamosals and the exoccipitals. The wide transverse expansion of the latter elements makes the base of the occiput very broad. The paroccipitals are likewise broad at the base and closely applied to the bulle, but the distal portion is slender and tapering. In these respects the inferior portion of the occiput is intermediate in structure and appearance between that of Hporeodon and that of Merychyus zygomaticus. As in the latter, the condyles project more posteriorly than in the John Day form. 130 THE MAMMALIA OF THE DEEP RIVER BEDS. The mandible is most like that of Hporeodon, but with some changes in the direction of Merychyus. Owing to the procumbency of the incisors and their alveoli, the outline of the chin, when viewed from the side, is seen to be strongly concave. In Hporeodon the posterior margin of the angle and ascending ramus is regularly rounded and projects far back of the condyle, while in Merychyus zygomaticus this margin is nearly straight and vertically directed and extends but little back of the condyle, from which it is separated by a notch. In Mesoreodon the shape of this region of the mandible is intermediate between these two extremes. The coronoid is short and slender, the sigmoid notch deep and widely open, and the condyle is much extended transversely. By a happy accident the hyoid (PI. III, Fig. 29) is preserved almost intact in one of the specimens and in its natural position. This apparatus in some respects differs from that of any known artiodactyl and agrees better with the hyoid of certain — perissodactyls. No doubt can exist as to the proper reference of the specimen, as is demonstrated by its connection with the skull, which was that of a large male, as indicated by the robust canines. The tympano-hyal is a short, stout, cylindrical bar, which is inserted into a de- pression upon the outer side of the auditory bulla. The stylohyal forms a long and broad (antero-posteriorly ) but thin and very much compressed bar, which expands at the proximal end, but this portion is fractured, so that its exact shape cannot be determined. Except for this proximal expansion, the bone is of almost uniform size throughout. The epihyal is well ossified and relatively longer than in the sheep ; it is narrower and somewhat thicker than the stylohyal and tapers distally. The cerato- hyal is also better developed than in the modern ruminants ; it is of a curious, paddle- like shape, slender and rounded where it joins the epihyal and expanding into a rounded blade posteriorly, where it is applied to the basihyal; it is not ankylosed with the latter. The basihyal is unlike that of any artiodactyl, with which I have been able to compare it, and much more resembles that of the horse. In shape it is nar- row, depressed and thin and curved backward; 7. e., with the concavity towards the front. Its great peculiarity, for an artiodactyl, consists in the presence of a glosso- hyal process, which is given off from the middle line of the hinder border. This process is much shorter proportionately and more curved backward than in the horse, and is compressed antero-posteriorly instead of laterally. I can find no other artiodactyl in which this process occurs. The thyrohyals are ankylosed with the basihyal, at which points they form slight, club-shaped swellings; they are slender, rounded, arched somewhat anteriorly and are of unusual relative length, being nearly as long as the stylohyals, though of altogether different shape. THE MAMMALIA OF THE DEEP RIVER BEDS. 131 Posterior to the hyoid apparatus, but not directly connected with it, the matrix contains a hollow, compressed semicylinder, or spout-shaped piece of bone (Fig. 29, T.c.), with exceedingly thin walls, which, strange to say, is unmistakably the ossified thyroid cartilage of the larynx. The position and, still more, the shape of this bone do not admit of the least doubt as to its nature, and this is one more added to the many peculiarities of this very peculiar family, though whether the character is con- fined to the present genus or is common to many other members of the group is not known as yet. Obviously, only by the rarest chance could such a fragile structure be preserved. Possibly the ossification of this cartilage is a sexual character, for, as already mentioned, the skull with which the specimen was found associated is very probably that of a male. The function of this bone was probably similar to that performed by the enormously inflated basibyal of the howling monkeys, and must have given to these animals most unusual powers of voice. So far as I can discover, such ossification is not known elsewhere among the Mammalia. There is no available material to compare the hyoid apparatus here described with that of the other genera of the family, since these bones are but rarely found in a fossil state. One specimen of Hporeodon from Oregon shows, however, that the stylohyal in this genus resembles in size and shape that of Mesoreodon. Measurements. MALE. FEMALE. M. M. Length of skull from summit of occiput to end of nose......-..-..-..+. .252 248 Length from occipital condyles to incisive alveoli.....-.........--..--+- 282 234 Distance from summit of inion to base of lower jaw (vertical).......... .180 .187 Length of cranium from summit of inion to postorbital margin ......... 1385 133 Length of face from postorbital margin to end of nose...............--- 126 126 Height of occiput........... 60-2. eee ee eee teens 085 078 RREACKIN OF TO eae AVG MAA Gaoesooobccnb0oc cn 000s 00000000 n00000000N00 .070 070 Breadth of skull at zygomatic arches............---- 21. e ee eee eens 142 142 Breadth of skull at postorbital arches. .........0..-2+ sees eee eee eens .110 oll? Breadth of face at upper first premolars.........-.----+++esseeeeee sree .049 OAT Breadth of face at last molars)... .- 00... seems: oe ee sae .080 085 Length of parictal crest.......--.. sees eee seen eee cee cent eter etnies 098 089 Length of frontals in median line..........-...see 5.2 eee eee etree eee 050 058 Length of nasals in median line.......-..+..+2 +++... eee ee etree eee 100 100 Breadth of face between frontal angular processes........--...-.++++++- 029 .033 iVerticalaudiameteroforbitiececerciter eee c reer Lei ret crt .030 .039 iiransversendiamelenseeee cence eect eile .030 -030 Height of lower jaw at coronoid process ......-.-++++++++s sess sees ees 096 eee Height of lower jaw at condyle.......---...ees sees eee t eee .090 095 Height of lower jaw back of last molar............s.e0ee eee seers eee .049 053 Height of lower jaw at third premolar.............-+.. +0. esse seer sees 031 034 132 THE MAMMALIA OF THE DEEP RIVER BEDS. MALE. FEMALE. M M Breadth of lower jaw obliquely back of last molar ................-.+-- 073 075 ILA CF YAMA VETS OF WOW PWioodoocpnoobovocs coe dbEbeodcoGuo00G000 D800 052 EISEN Of QMIGTO MAKE onconcondnocqondoooconsoado boo DODSDOODODDOGNDO 085 032 Breadthpoteantenionmanesiencrt tree eect renee kteerierts 026 032 Distance between the supraorbital foramina ...........-.00+-- ee ee ee eeee .016 024. Distance between the infraorbital foramina..... Eco ona name cnTneCnOOKG 056 060 IeMANO, Ol Ihe! SORE cooogceDDbGoDDDODDG0DoeD GDh oaDGDDO0DDD0000C0GD000. oocL 126 Width of hard palate between first molars.............-...---.---+ 0-205 were -045 Stylonyall lens thence ieee ieee rei riet i ren t-te fcr tere 2.050 DMA, WEBI Goconssssedosoccconsonsbovecgdoob00ndds0d80000000000000 015 Cermi@inyall, SHEN cocco¢ccoscccccca canons oedogobOODnODOGEeED UOC DOD CUNC 019 BRAM, WENN. occ oogs0dcocop0ud0oG00000 HoaNoWOCU Nooo IuOn0000000 018 AMG ARO OAL, NOAM, oo ooovoonscecsasd coud Dc90bAb ONC DOGOUNODUSOGdUODOOROON .040 Tin proG| Garbleee, INGEN soosoooc sovausdeba0gncococuoNonoGD0es0000000 -016 The Brain. Having no brain-cast of the John Day genus Hporeodon prepared, IT am unable to compare that of Mesoreodon with it, and must therefore take the White River Oreodon as a standard. In the latter genus there is a considerable degree of individual variation, both in the shape of the hemispheres and in the num- ber and extent of their convolutions. To some of these types the brain of Mesoreo- don presents a much closer resemblance than to others, but is somewhat more advanced and modernized than any of them. This advance is noticeable in the gen- eral form of the hemispheres, which, though not broader behind than in some speci- mens of Oreodon, are much more so anteriorly, and thus the whole brain is fuller, more rounded and tapers less forward. The hemispheres have also slightly increased in vertical diameter, so that they are no longer exceeded in this dimension by the height of the cerebellum and medulla oblongata. Posteriorly, the two halves of the cerebrum are brought closer together and reach the cerebellum in the middle line, not gaping so as to expose part of the optic lobes, as is the case in the White River genus. They do not, however, appear to overlap the lateral lobes any more exten- sively than in that form. The sulci are very nearly the same as occur in some specimens of Oreodon, but they pursue a slightly more sinuous course, which gives an appearance of richer con- volutions. The dorsal surface displays (1) a short and straight lateral fissure, which does not connect anteriorly with the suprasylvian, as is sometimes the case in the White River genus. As regards the latter, Krueg (No. 22) regards this sulcus as the splenial, which by an extreme degree of “‘ supination.” is exposed upon the dorsal surface of the cerebrum, as in many of the small artiodactyls now living. But as this fissure does not extend to the medial surface of the hemisphere, this interpreta- THE MAMMALIA OF THE DEEP RIVER BEDS. 133 tion does not seem probable. (2) There is a longer and more curved suprasylvian fissure, which is connected anteriorly with the coronal fissure by means of a short and faintly marked ansate sulcus. This connection is also found in some specimens of Oreodon. (3) The coronal sulcus consists of two portions; the anterior is longer and curves downward and outward, while the posterior is shorter and more obscurely marked and converges towards the middle line in a way that suggests the crucial sulcus of the Carnivora. The lateral view shows, in addition to these fissures, a short and nearly horizontal sylvian sulcus and a presylvian which has a more nearly vertical course than in Oreodon. The sylvian fissure appears to be connected with the fissura rhinalis, though in this region the sulci are very obscure and diffi- cult to interpret. Indications of a posterior suprasylvian sulcus are also to be observed. The character of the cerebral sulci is, it is obvious, very much the same as that which occurs among the smaller and more primitive forms of existing ruminants, and these, as Krueg has shown, agree closely in fundamental plan with the Carnivora. As there is every reason to believe that the Oreodontide are connected with the Pecora only through very ancient forms, in which the hemispheres were either smooth or but very little convoluted, this resemblance must be chiefly ascribed to parallelism of development. Still more obviously is this the case with regard to the likeness between these artiodactyls and the Carnivora. In both of the brain-casts of Mesoreodon the olfactory lobes are broken away, but it is plain that they were not at all overlapped by the cerebrum. The cerebellum is very much as in the older White River type; its posterior face rises nearly verti- cally from the medulla; the vermis is large and prominent and the lateral lobes are broad. In neither of the specimens is the cerebellum sufficiently well preserved to display the details of the convolutions. The Vertebral Column. The atlas is rather more like that of the true ruminants than is that of Hporeodon. This is due principally to the more uniform width of the transverse processes and their continuation into short spines behind the surfaces for the axis, from which they are separated by decided notches. This prolongation of the transverse processes is, however, much less marked than in the Pecora. On the other hand, the processes are more widely expanded laterally than in Hporeodon, which is a departure from the ruminant type. The anterior extension of the trans- verse process has, as in the earlier genera of the family, converted the atlanteo-diapo- physeal notch into a foramen, but there is no perforation for the vertebrarterial canal. The anterior cotyli for the occipital condyles are deep and are more distinctly sepa- rated at their inferior borders than in Hporeodon and the neural spine is larger and 134 THE MAMMALIA OF THE DEEP RIVER BEDS. more rugose. The posterior articular surfaces for the axis are larger in both dimen- sions than in the latter genus, but especially in width. The axis is very different from that of the Oregon genus. The atlanteal facets are broader and higher, descending more below the level of the centrum and separated by a more decided medio-inferior notch, but not, as is the case in Hporeodon, divided above by deep notches from the bases of the pedicels of the neural arch. The odon- toid process is wider, more depressed and spout-like, with more elevated margins and with the articular surface for the inferior arch of the atlas rising higher upon its sides. ‘The odontoid process is thus in an analogous stage towards the assumption of the spout-like character as it is in Profolabis among the camels and Miohippus among the horses. The transverse processes are longer and heavier than in Hporeo- don. The neural spine is of very different shape from that of the latter, a change which is chiefly brought about by an elevation of the anterior portion, so that it forms a large hatchet-like plate, quite different from the spine found in the other members of the subfamily and more like that of Agriochwrus. The postzygapo- physes are more horizontal in position than in the Oregon genus, presenting more directly downward and less obliquely outward and backward. The pedicels of the neural arch are not perforated for the second pair of spinal nerves. The remaining cervicals are rather short, with slightly opisthoccelous centra, which are keeled on the inferior side. The transverse process and pleurapophyseal plate are well developed and the latter reaches great size on the sixth vertebra. The neural spine is a mere ridge on the third, fourth and fifth cervicals, on the sixth it is much higher, slender and inclined forward, while on the seventh it is still higher and heavier. In proportion to the size of the head, the neck is of about the same length as in Hporeodon, but the vertebre are more heavily built. The thoracic vertebra are not different in any very important respect from those of the Oregon genus, except for the better development of the spines. On the first of the series the spine is considerably higher and thicker than on the last cervical, but is much surpassed in both respects by the spine of the second, in which this pro- cess reaches almost bovine proportions. ‘The other anterior thoracic vertebre have broad, compressed spines, though none of the specimens are sufficiently complete to allow a determination of the length of these processes. The remaining regions of the vertebral column are represented by numbers of isolated centra, from which the processes have been broken away, and which there- fore do not require any detailed description. So far as they go, these bones differ but little from the corresponding vertebrae of Hporeodon. The anterior ribs are rather short, broad and compressed, and of triangular THE MAMMALIA OF THE DEEP RIVER BEDS. 135 section; posteriorly, they become longer, much more slender and of more rounded section. Nothing is known of the sternum. The Fore Limb displays some characters of unexpected interest. The scapula varies considerably in the different specimens, some of which variations would appear to be of a sexual nature. The best preserved shoulder-blade (PI. IV, Fig. 33) is one which belonged, as I believe, to a female, being associated with a skull in which the small size of the canines and the caniniform first lower premolar is very striking. In this specimen the coracoid and suprascapular borders are broken in such a way as to prevent an accurate determination of the outline of the bone, but the spine, neck, and most of the postscapular fossa are in good condition. The glenoid cavity is small, shallow, and of nearly circular shape, the antero-posterior diameter but slightly exceeding the transverse. The coracoid is large, especially in the vertical dimension, but is not clearly demarcated from the neck of the scapula and displays but little rugosity ; hence it is not conspicuous when viewed from the outer side. The neck of the scapula is high in the vertical direction, narrow and contracted, and the rugose lines for muscular attachment are but faintly marked. Above the neck, the glenoid bor- der extends obliquely upward and backward, enclosing with the spine a narrow, triangular postscapular fossa. This border is considerably thickened and its external margin is elevated, making the fossa quite concave antero-posteriorly. The spine is very high, and for most of its length curved backward, so as to make the anterior surface convex and the posterior concave. Its free margin is flattened and gradually becomes wider inferiorly to the point where it sends out a distinct metacromial pro- cess. No other genus of the family has yet been found in which a metacromion occurs. Beneath this process the curvature of the spine is reversed, the posterior surface now being convex and the anterior concave, and the acromion projects for- ward as well as downward. The length of the acromion cannot be definitely stated, as its tip is broken away, but obviously it could not have descended nearly to the level of the glenoid cavity. This spine is of a very exceptional character for an artiodactyl. In Oreodon there is no metacromion, the spine is lower and not recurved and descends more nearly to the level of the glenoid cavity. In Hporeodon there is likewise no meta- cromion, but the spine is very high and curved in very much the same fashion as in Mesoreodon. The second specimen (Pl. IV, Fig. 34) has lost the spine but preserved the entire outline of the scapula, which is represented either by bone or by the impression of it in the matrix. There is good reason to believe that this specimen should be referred A. P. S.—VOL. XVIII. R. 136 THE MAMMALIA OF THE DEEP RIVER BEDS. to the same individual as a large male skull which was found in the same locality, but on this point I cannot speak with entire confidence. The glenoid cavity is some- what larger and more oval in shape than in the specimen first described, the antero- posterior diameter distinctly exceeding the transverse; the neck is broader, heavier and less contracted, and has well-marked rugose lines for the attachment of muscles. On the other hand, the glenoid border is less elevated and the postscapular fossa is, in consequence, less concave. The coracoid border curves forward much more decidedly than it does in the scapula of Oreodon, making the proximal portion of the blade relatively much wider than in that genus. The suprascapular border is gently arched, but nevertheless forms nearly a right angle with the glenoid border. The spine is almost median in position, giving pre- and postscapular fosse of nearly equal size, but of different shape, owing to the different course taken by the coracoid and glenoid borders. The block of matrix which contains the scapula just described, holds also the last four cervical and first four thoracic vertebra, with their ribs attached, and the proximal half of the humerus. ‘The same block contains also a small bone (Fig. 34, cl) which is removed but a short distance from the coracoid process of the scapula, and runs forward and inward, overlapping the first rib and the transverse process of the seventh cervical vertebra. Zhis bone I regard as a rudimentary clavicle. Nat- urally, such identification will be received with much doubt, and I was at first very skeptical about it myself. It is certainly most unexpected to find this element in an ungulate so far advanced in differentiation and so high in the geological scale as the middle Miocene, while it has not yet been detected in the Condylarthra of the lower Eocene. Nevertheless, in spite of the a@ prior? improbability of the occur- rence of the clavicle in a Miocene artiodacty], there seems to be but little doubt that such is actually the fact. In the first place, the position taken by the bone in ques- tion is such as a clavicle would occupy if it were present. There is a slight vertical displacement of the whole fore limb and shoulder girdle, but otherwise the bones of all the surrounding parts—vertebre, ribs, scapula and humerus—are in their natural position almost as perfectly as in a living animal. (2) There is no other bone in this skeleton with which this one can be identified ; it is much too slender to be a part even of the smallest rib, and its shape is quite different from that of any of the elements of the hyoid apparatus. Fortunately, we already possess the latter belong- ing to (presumably) the same individual and can definitely state that the bone in question cannot be referred to it. (8) The shape is that which we should expect to find in a rudimentary clavicle; it is slightly arched downward, is very slender and of rounded section, with an inferior keel, which is best marked in the middle and dies THE MAMMALIA OF THE DEEP RIVER BEDS. 137 away towards the ends. (4) The unusual development of the spine of the scapula is an indication that the clavicle had not been entirely lost. (5) Wineza’s observa- tion (No. 36) that a transitory rudiment of a bony clavicle is developed in the sheep, points to the conclusion that this element has not been eliminated from the artiodac- tyls for so long a period as has been generally supposed. (6) Admitting that the structure under discussion really represents the clavicle, its very small size and loose attachments (for it is in contact with neither the scapula nor the sternum) will ex- plain why it has not yet been observed in the more ancient forms of ungulates. Only by the rarest chance could such a bone be preserved in its natural position. Unless, therefore, we are prepared to assume that a single bone of some small animal has become accidentally entangled with this skeleton of Mesoreodon, and in such a way as to exactly simulate the position of the collar bone, which is certainly highly improbable, there would seem to be no escape from the conclusion that the clavicle was present in this genus. However, other specimens will be required before we can be entirely satisfied on this point. In this connection it may be noted that the simpler and less developed scapular spine of Oreodon would lead us to infer the absence of a clavicle in that genus. But, assuming this to be the case, we cannot yet determine the significance of the fact since so little is known of the skeleton of those White River species in which the tympanic bullee are inflated and which are presumably the ancestors of Zporeodon. In the absence of knowledge on this point we cannot tell whether the supposed clavicle of Mesoreodon should be regarded as a persistent rudiment or as a case of reversion and the reacquisition of a lost structure. The former alternative would certainly seem to be more probable, and, if it is true, it may serve to explain the very general differ- ence between artiodactyls and perissodactyls with regard to the development of the acromion. As is well known, this structure is in nearly all artiodactyls large and - prominent, while in even the Eocene perissodactyls the acromion is absent. If we may assume that the clavicle persisted lorger in the former group than in the latter, this difference would be accounted for. The humerus (Pl. IY, Fig. 34; Pl. V, Fig. 37) is in general very similar to that of Lporeodon, but has a decidedly stouter shaft ; the head is more convex and presents -much more posteriorly, less exclusively in the proximal direction. The external tuberosity is of a different shape, its extremities being less produced as overhanging hooks; the internal tuberosity is also less developed, and in consequence the bicipital groove is not so deep. The length of the shaft is about the same as in Hporeodon, but the diameter, both transversely and antero-posteriorly, is much greater. The distal end is of the shape which is characteristic and constant throughout the family. 138 THE MAMMALIA OF THE DEEP RIVER BEDS. This shape is already well shown in the White River species and is marked by its relatively great transverse breadth, prominent internal epicondyle, broad, rounded intercondylar ridge, which is nearly median in position, and the nearly equal width of the external and internal divisions of the trochlea. The anconeal fossa is relatively higher and narrower than in Lporeodon. Some differences, which are probably of a sexual nature, are to be observed in the proximal end of the humerus in different specimens. In those which are asso- ciated with the skulls marked by small canines, and therefore presumably female, the head is more hemispherical and markedly shorter in the antero-posterior direction ; the external tuberosity extends less completely across the anterior face of the bone and the bicipital groove is wider. Possibly, however, these distinctions are specific rather than sexual. The ulna and radius (Pl. V, Fig. 38) are massively constructed; they are en- tirely unreduced and show no tendency to codssify at any point. ‘The radio-ulnar arch is very long, extending from a short distance below the head to the distal ex- pansion; this is not visible in the anterior view. The radius has the form of head which is characteristic of the family and therefore requires no detailed description. The shaft is not so broad and antero-posteriorly compressed as in Hporeodon, but more rounded and cylindrical in the middle, reverting thus in some degree to the condition found in Oreodon. 'The distal end is more expanded transversely than in Eporeodon, though in this respect there is considerable variation. The scaphoid sur- face is peculiar in the deep groove on its ulnar border, This is already indicated in the Oregon genus, but to a much less conspicuous degree, especially in the breadth and depth of the groove behind. A similar groove appears in Merychyus. The lunar facet is like that of HZporeodon, but is less closely connected with the scaphoid facet. The ulna is quite unreduced and has a very heavy shaft, which almost equals that of the radius in antero-posterior thickness and exceeds it in width. The olecra- non is high and massive. The distal end is excavated to receive the expansion of the radius and carries a facet for the cuneiform, which is narrow antero-posteriorly but broad transversely. This ulna differs but slightly and in no important respect from that of Hporeodon. _ The manus (Pl. V, Figs. 39, 40) presents some features of much interest, as here we find most strongly emphasized the tendency towards Merychyus which is more obscurely indicated in the structure of the skull and teeth. In the carpus the scaphoid has undergone some noteworthy changes as compared with that of the more ancient genera. It is increased in size, especially in breadth; the radial surface is, as usual, concaye behind and conyex in front, but rises more towards the ulnar bor- THE MAMMALIA OF THE DEEP RIVER BEDS. 139 der to enter the groove on the radius already mentioned. The ulnar side of the scaphoid is concave and is chiefly oceupied by the large inferior facet for the lunar. The distal side is taken up by two facets, those for the magnum and trapezoid respec- tively. _The magnum facet is the larger of the two and is deeply excavated behind, but descends abruptly in front. The lunar is both high and broad; its radial surface is saddle-shaped, concave from side to side and convex fore and aft, broad in front, much contracted and tapering behind. The radial side carries two facets for the scaphoid, the upper one small and nearly piane, the lower very large and convex and separated from the magnum surface by a scarcely perceptible ridge. The latter facet is almost entirely lateral, except on the palmar border, where it is reflected under- neath so as to be partly distal. The unciform facet is concave and obliquely placed ; it forms with the magnum surface a sharp beak, which is wedged in between the magnum and the unciform and extends nearly to the third metacarpal. The cuneiform is broad and low and has a less dorso-palmar diameter than the other proximal carpals. The ulnar surface is a narrow groove which is reflected down upon the external side of the bone, and the pisiform facet forms a broad band upon the palmar side which is continuous with the ulnar facet. Distally, the cuncei- form displays a simply concave facet for the unciform. The pisiform is intermediate in character between the condition found in the earlier and that in the later genera of the family, being more expanded at the free end than in the former, less so than in the latter. The proximal end is much contracted and bears a single rounded articular surface, part of which is for the ulna and part for the cuneiform. The trapezium is a small nodular bone which has but two facets, one for the trapezoid and, at an obtuse angle with this, another for the second metacarpal. This species and Merycocherus montanus are the only members of the family in which I have succeeded in obtaining the trapezium, though the facets on the neighboring bones leave no room for doubt as to its presence in the other genera as well. This carpal enables us to state with entire confidence that in Mesoreodon the pollex is not represented even by a rudiment. The trapezoid isa large bone both vertically and transversely, but it has no great antero-posterior depth; proximally, it bears a large and simply convex facet for the scaphoid and its radial side is occupied by a concave surface for the trapezium. Distally, there are two facets, one of which is large and slightly concave, the other smal], plane and inclined at an open angle to the first ; the former is the surface for the second metacarpal and the latter for the third. The magnum is a very characteristic bone, resembling strongly that of Mery- chyus and Merycocherus, though its peculiarities are not carried to such an extreme. Seen from the front, the magnum appears to be smaller than the trapezoid, and is 140 THE MAMMALIA OF THE DEEP RIVER BEDS. both lower and narrower, but its upper surface rises rapidly towards the palmar side, forming the “head.” Behind the anterior face the bone is deeply constricted by two concave facets, one on the radial side for the trapezoid and the other on the ulnar side for the lunar. The trapezoid and magnum are very closely interlocked and form a continuous saddle-shaped surface for the scaphoid, which in appearance resembles the astragalar trochlea of a carnivore. The magnum, including even the head, is entirely beneath the scaphoid, the opposite condition to that of the horse, in which the head is entirely underneath the lunar, though the scaphoid rests upon the anterior portion. The lunar surface is deeply concave (though less so than in Merychyus and Merycocherus, in which it describes a semicircle) and almost entirely lateral in posi- tion, but on the palmar side is a small, shelf-like projection which extends somewhat beneath the lunar. This gradual displacement of the magnum towards the radial side of the carpus is already indicated in the oldest known genus of the family, Pro- toreodon, and is more decidedly marked in Oreodon and Hporeodon, though even in the latter it has by no means been carried to the same extent as in Mescreodon, in which it attains almost the extreme position found in Merychyus and Merycocherus. The contact between the magnum and the unciform is very slight and nearly or quite limited to the posterior or palmar margin, the two bones being separated by the long beak of the lunar and the strong process which the third metacarpal sends obliquely upward and outward to abut against the unciform. Distally, the magnum bears a single saddle-shaped facet for me. iii, which is reflected upward more upon the ulnar than on the radial side. This facet is elongate and quite deeply concave in the dorso- palmar direction, contracting to a point behind, narrow and very convex transversely. There is no facet for me. ii, that bone being excluded from contact with the magnum by the connection of me. iii with the trapezoid. The posterior hook of the magnum is short, curved, blunt, depressed and curved towards the radial side. The unciform is high and broad, with its proximal portion contracting posteriorly. The upper sur- face bears an oblique facet for the lunar, which rests almost entirely upon the unci- form, and somewhat larger convex facet for the cuneiform. The metacarpal surfaces form a nearly continuous curve. On the radial side, though confined to the dorsal half of the bone, is a large oblique facet for the projection from me. iii; distally, there is a larger facet for me. iv and a smaller one for me. v; the latter surface is reflected up upon the ulnar side of the unciform. The metacarpals are four in number and in their proportions very similar to those of the older genera, Oreodon and Hporeodon, though differing in some important respects from the metacarpals of those genera in their mode of articulation with the carpus, which is like that of Merycocherus and Merychyus in being of the “ adaptive” THE MAMMALIA OF THE DEEP RIVER BEDS. 141 type. In the second species of Mesoreodon, M. intermedius, to be described hereafter, the proportions of the metacarpals are more like those of Merychyus. The second metacarpal is short and slender, with a trihedral recurved shaft; the head is rather broad and bears a triangular, nearly plane facet for the trapezoid and a postero-internal one for the trapezium. This bone does not reach the magnum, which constitutes an important difference from the manus of Hporeodon, in which, as in the more ancient genera, me. ii is in contact with the magnum. The third meta- carpal is likewise different from that of the last-named genus; the head is expanded and deeply concave transversely and convex antero-posteriorly ; on the radial side is a very small oblique facet for the trapezoid and on the ulnar side a very large one for the unciform. Both of these surfaces are confined to the anterior half of the meta- carpal. Beneath the unciform projection on the head of me. iii the bone is excavated to receive the head of me. iv ; posterior to this, and separated from it by a deep sulcus, is a second facet placed on a projection, which extends towards the ulnar side. This facet is somewhat oblique and extends beneath the head of me. iv and the two bones are thus interlocked in a very complex and perfect manner. The same arrangement is indicated in Oreodon, but in this genus the posterior facet is much less conspicu- ously developed. In the present species, the shafts of me. iii and me. iv are quite heavy and not very long, in which respect they differ very markedly from those of the second species, MW. intermedius. In Oreodon, me. iii not only rises above me. iv but also extends below it distally, while in Mesoreodon, as in Merychyus, me. iv extends slightly below the end of me. iii, though it is considerably the shorter of the two. Proportionately, me. iv is little, if any, longer in Wesoreodon than in the White River genus, the different disposition of the metacarpals being due to the enlarged unciform process of me. iii, the greater relative height of the unciform, and the con- sequent downward displacement of the head of me. iy. The fifth metacarpal is quite different from that of Oreodon in having a narrower but deeper head, with the shaft broader proximally, expanding less distally and being more strongly recurved. In length and thickness it is the counterpart of me. ii, whereas in the White River genus the latter is decidedly stouter. On all of the metacarpals the distal carinz are much better developed than in Oreodon and are plainly visible when the manus is viewed from the front. The phalanges of the proximal row are like those of the earlier genera of the family, except that they are relatively shorter. Those of the second row are notably shortened and broadened ; the distal trochlea is wider and not reflected so far upon the dorsal side of the bone, but is more prominent upon the palmar side. The ungual phalanges of this species are very peculiar and different from those of any other 142 THE MAMMALIA OF THE DEEP RIVER BEDS. member of the family. They are wide proximally and taper distally to a point, which is more acute than in the other genera; in the median pair (digits iii and iv) the approximate borders are straight and the opposed borders curve distally towards the median line. Proximally, the ungual has considerable vertical depth, but this diminishes rapidly towards the distal point, and the dorsal surface is very convex and strongly arched from side to side. The unguals of the lateral digits are like those of the median pair, except for their very much smaller size. In brief, the ungual phalanx is shaped like the half of a slender and somewhat irregular cone. Measurements. MALE. FEMALE. M. M. \ Scapulaspheic hiteereepeeeeec eee lccrcrteriiieeCei ee eeeerercr ner tien: 138 2.140 SOUR, SEMI TWIN coopccoonnccoocboDNsudUaDFCDaDaNdOUDSSGDGOODbRG 088 2.090 Seno, Win OF iGOKooonsascaccgb0c 0D po0AnoHOboDD0ODDONbDd0000000000 028 021 Glenoid cavity, antero-posterior diameter..................e0eeescecceee 027 -022 Glenoiducavity, itransverses diameter -.-)-1c1cicinsls(leleslelele elelsiei-ieisielslelelenes e/iel> 025 022 TE(VINETTE, NOVI, sogqcco ccs cogs ooNnOODCODo0 GD OOS OOODO DN BDYGOOCOBDOONOOO 608 151 Humerus, antero-posterior diameter of proximal end..................-- .056 049 Humerus, transverse diameter of proximal end.......................-- .042 .039 umerussawidthwofadistalmenderemesererccee cet oet een rice eenieeeenne -039 HAGE, WDA cecocoscrascoacocacdddonnasooodds nan sscdcnuepnooeOObdS 2.027 Beare TRCN, TOMES oocasnscceboacdedooncanbeesHSenono UNOS HD00ND ON GS DHONOONNS odDe 137 Radiussamidthwoteproximalmen denne silaee ci iieiiececer tee erierictir merit 027 Radiussswidthwofadistalsendererer rere eee eretncicccr eccrine .029 IRAchinNE, THGlih OF Ghats in WNCoocc0cccccoosn00ccd00Dso5cHEHU0000000G0 acCD .014 Lilien None adeen tines Sanam ana mascara ad sete aac seb a anay asae edo .180 Ulna, height of olecranon from sigmoid notch.......c0..sceseee- eee ees wees -030 Winasbreadthpofashattwimemiddlemerrcrtcrecimereeeeerere citer eect rete imeaciac .016 UWinas toreadthy of: tdistal\en'dicwleve-ccsortt ie roreislsicic tere cvele sistaiereiovee (oleretaveveticierstersis elites -014 Scaphoidvshel chiteriserececer rere eerie ete eierriielertercierr creer irer relat -016 Scaphoids widthwot proximval endherrisertetacce ieee eleieeiars 012 IIMA, INSAN oo sbocodsaddo basse oDcesosupesdabubdsoGDDDDOCSOuSAGOUDDODO O17 Imunarawidthwofeproximalmendtermeercerec ccc ceeremertsicciiir re ettelere -012 Pisifonmeseleny tlie ey etic ielcee ate eiciete tedster ieletsteketeieieieietctatetele 023 Metacaxpall i, lengthe = ccc cic posntendsadocoDnSocDeeDeaKG000N000 054 Metacarpal iii, length (not including unciform process)................- 065 Metacarpall ii, wid thtof pproximalendyjrtei)ote.o- lelelelelelelsiel-lelalasiereietelatelersierele 016 MetacarpalliilswadthyofedistalitenGecreiertetleter tle latalololelotteteleisiefoferattoletaleL-tereketets 014 Metacarpal ili, width of shaft below head..........................-..-- -013 Metacarpal iv, length .-2........-.2.0-----< Soo DD odaCoRbsdCaOoCDOGGOCO0 -O61 Metacarpalliva widthmotesproximal endlyrrrel jects ler leri-lelcieeteielterichasinreciitetete .013 Metacarpalinveswid thwyofadistalmenchyerridelleretiiieriertite ricci terre irtetlerers -012 Metacarpal iv, width of shaft below head................-.-.020000+-00> 012 Meta carpalliavailen Dit liseverstertat-!tapereteletetterltelafeveralstelolelatelotedctevetetst tel steleie tetany tercters -052 Hirstyphalanx ofthindedisit;eleme herent case clr eect eee .021 THE MAMMALIA OF THE DEEP RIVER BEDS. 143 Second phalanx of third digit, length..................-.....-se000e--> 013 Dhirdsphalanxsofithindyd iste] eme tyecrrtetetereiletetsteteleteretersterey=ielelereteyelelelelelei= 017 Third phalanx of third digit, width of proximal end.................... 009 The Hind Limb presents fewer peculiarities than the fore limb, and, except in a few details, is very similar to that of Hporeodon. The pelvis (Pl. VI, Figs. 46, 47) is very like that of the Oregon form but with some minor differences. The ilium has a shorter peduncle which expands more abruptly into a wider plate. The latter is less strongly everted, especially at the antero-inferior angle, which is less prolonged. The iliac surface is broader and more rounded, the acetabular border less prominent, and the pubic border more so. The pit for the rectus femoris muscle is larger but not so deep. The acetabulum is much larger and relatively shallower and the artic- ular surface is more reduced by the very large sulcus for the round ligament. The ischium is more twisted upon itself, so that the posterior end is much more everted and depressed. The crest above the acetabulum descends more abruptly in front and dies away behind without forming an ischiadic notch. The pubis in its free portion is short and stout and the symphysis, in which the ischium shares, is elongate. The femur shows analogous differences from that of Hyoreodon. The head is distinctly larger and more sessile and has an unusually anterior position; it does not rise so far above the bridge connecting it with the great trochanter. This bridge is more thickened in the antero-posterior dimension and the great trochanter is larger and more massive. The shaft is heavier and more arched forward and, distally, is both broader and deeper. The external linea aspera appears to be less conspicu- ously marked and the pit for the plantaris muscle shallower. The condyles have a greater vertical diameter but do not present so strongly backward. The rotular troch- lea is wider; its margins are of equal height and more compressed. The tibia is very much alike in the two genera. In Mesoreodon the external surface for the femoral condyle is broader and the cnemial crest more massive and rugose; the shaft is slightly heavier and the distal end rather more expanded, both transversely and from before backward. The grooves for the astragalus are wider and the intercondylar ridge broader and lower and not forming, as in Merychyus, a posterior tongue. The sulcus which invades the external astragalar groove is larger and deeper, and the external groove is a little wider relatively to the internal than in Eporeodon. An important change which is already indicated in the Oregon genus is carried farther in Mesoreodon, viz., the presence of a distal facet for the fibula, showing that the latter has extended slightly beneath the tibia. I cannot ascertain the condition of the older species of Merychyus in this respect, but in a specimen belonging probably to MW. elegans, from the upper Loup Fork, there is no fibular facet A. P. 8.— VOL. XVII. &. 144 THE MAMMALIA OF THE DEEP RIVER BEDS. on the distal surface of the tibia and the sui face on the external side is much larger and more deeply impressed. The distal end of the fibula is preserved in one specimen; it forms a very large and heavy external malleolus which is especially expanded antero-posteriorly ; on the distal side is a very long facet for the calcaneum, with the outer border somewhat elevated. On its inner surface the malleolus displays a projection which extends beneath the facet already described on the distal aspect of the tibia. The shaft was obviously reduced to the most slender RS a though we cannot determine whether it was interrupted or entire. The pes (PI. V, Figs. 41, 42) requires but a brief description, as it departs very little from the type common to nearly all the members of the family, there being less variation in the structure of the hind foot than of the front. The caleaneum has a long and stout tuber, with nearly parallel borders, and thickens at the free end into a heavy knob; the fibular facet is low but very long, in which respect it differs strik- ingly from that of Merycoidodon. The distal astragalar facet is very long, but, as in nearly all the genera of this family, the sustentaculum projects but very little. The cuboidal surface is narrow but long, measured from the dorsal to the plantar edge. The astragalus is low and broad; the outer proximal condyle exceeds the inner considerably in size, but less than in Oreodon, and is invaded by a larger sulcus; it has a more thickened and gently rounded external border than in that genus and the intercondylar groove is wider and less angulate. The navicular surface is very differ- ent from that. of the White River genus, extending higher up upon the anterior face of the bone and having decidedly greater dorso-plantar thickness. Corresponding to the structure of the caleaneum, the sustentacular surface is long, narrow, and not connected with other facets. The cuboid, compared with that of the earlier genera, has increased in relative height, which gives it an appearance quite like that of Mery- chyus, though it has not attained such an extreme; the calcaneal facet differs from that of Merychyus and resembles that of Oveodon in being incised lower down upon the anterior face of the cuboid. The navicular has undergone no important changes, unless the greater elevation of its antero-external border be so regarded; the pos- terior hook is very long. The cuneiforms are very much as in the older genera. As in all the members of the family, the ecto- and mesocuneiforms are codssified, the com- pound bone differing from that of Merychyus (at least of such species as JV. elegans and M. arenarum) in its much greater proportional transverse width. The metatarsals resemble those of Oreodon, but some changes may be observed which point in the direction of Merychyus. Thus, the metatarsals are straighter, less THE MAMMALIA OF THE DEEP RIVER BEDS. 145 arched forward and more slender in proportion to their length; the distal carine are more prominent and more extended upon the anterior face of the trochlea. The phalanges (Pl. V, Fig. 42) of the pes are like those of the manus, except for their greater size, and the curious, trowel-shaped unguals have their peculiarities somewhat emphasized. Measurements. Diameter of-acetabulum,, toresand!attse-ee eee eee eee eee eee Bek een eee meee ey te 029 ID MAG OF ACAI MIMI, WEA oc cooccpoosoc Des cousHoadasonBooADasennoDoooOOOssSUOOOESS eee 028 Width of pelvic opening at pubes................. Sey ae foes Spon Pret Rome ea nant atereNS aster sy ee reer ne .059 Nannie, WANG Nos ocodcpouscsncgcced cos GIAO soDT DDG CAD CDOs aDHAdESOODODDIODSOCOGRENSIOONNBNN .190 » Femur, width of proximal end............ SHiSde Seas eons hases ads obep cascada S poab ep couOsecaD .050 Hemunidiametenotehesdetoresandwattaeemeerrerreeer nice eeeieer etic iti eter terrain) 025 MMI, HOAs oc ca0ananss seco gdaueaocaN cdo SET Us oesG DSS DD DOOD DUGOBOUaDORUE.OOUDSDOOSOSOS ATT Uo, OAC OF jORORTN GME sosoaccacssocoonoqa pons ocaso0socd oe sNOUE soo ROOD SORODNDAACO 044 uRibiassbreadthyofadistalaen dsereers-eeereee eee ceeCrencerr eatin teeta feet terra 028 ARETE, UGB Ns ooocadoossbossans a0 bbd8CouasdesposuagHDDUde bod duoDe SO Oo AgUHORAGOSODSDESE 035 Astragalus, width of proximal trochlea.................-...- +222: s 2-2 ee eee eee .020 C@alcaneum»plente (hesee er eeen eee sane eee eee eee eee Gr eee ae mente rect Tittle .064 ClMlowiel, VARs cos copa senoabosoncooquovadsoouSddoocoEDdsadEonAcaSoOnmEgcaCHoAbDSyocCODOONGCNC .020 Wistar thy, WM EWN Gs cooses pan edocosscoodconcoosoe sda cdenEdanS OOhodsos0qqgCedSCODNNGRRC 075 WIGENEMEEIL Thy, WACHIN Ci? jonah OnGlosoaccqnssonogaccsccsudubo00o0adbo0G0 Jadu DoD Dee SObeo0uRG 018 First phalanx of third digit, length .........----- 2-22.22 eee 2 eee cee ee eee cette .022 First phalanx of third digit, width of proximal end ...........-.-...-.. sees eee e tees eee 013 Second phalanx of third digit, length.............--- 2 see eeee eee eee eee etnies 012 Third phalanx of third digit, length............---.ceeeeee esse eect eee ete cees 016 Third phalanx of third digit, width of proximal end.........-.----.--+-.++-++seeseeeee eee 013 MESOREODON INTERMEDIUS Scott. Amer. Naturalist, 1893, p. 661. This species is represented by foot-bones belonging to two different individuals, which are so different in their proportions from those of the foregoing. species that they must be referred to another animal. The only skull which can with any proba- bility be regarded as belonging to MZ intermedius is a small one belonging to a very immature individual and therefore of little value for systematic purposes. The milk teeth are still in place, the permanent canines just beginning to appear, and the first permanent upper molar already protruded. This latter tooth is somewhat like that of Merychyus in the shape of the external crescents and in the rapid narrowing of the valleys towards the base of the crown. The inner crescents, however, are not like those of the Loup Fork type. One of the typical specimens (PI. V, Fig. 43) consists of the third metacarpal entire and the proximal three-quarters of the fourth. As compared with those of 146 THE MAMMALIA OF THE DEEP RIVER BEDS. the preceding species, these bones are decidedly longer and more slender, with the carinee of the distal trochlesze more extended anteriorly. In their proportions and in their carpal facets these bones closely approximate these of Merychyus, though the trapezoid facet of me. iii is smaller. ; The second specimen (PI. V, Fig. 44) consists of isolated phalanges. ‘Those of the proximal row are much more slender and arched forward than in MZ. chelonya. The unguals are extremely like those of Merychyus, but are somewhat more obtusely pointed. Measurements. M. Meta carpal biiilen pth yer ja-tvste cies erate encttomir fete es Petts eter ea eerNe eLetters Beers cise .069 © Metacarpalmiiipswidthwofiproximaliends-eeeesneee eee eee ericrieccriciocciserienee ete 015 Metacarpalliijwidthwofedistaliendtaa--eereaanaeenaececiienerineceoe ocecicieecareicrricion .014 Metacanpalmiawidthwofashattibelowsheadienestereeee eterno terror esce tert .O11 Metacanpaleivanwidthkofsproximaluend@rrreceecaceceeee eer eeeeeee eerie eee 013 Metacarpalpiveawidthwofeshaftubelowahcademsrerrecemccetoeciaeic iti ricer ite nie teirererr rani OLL Hirstephalanxaofsuhirdudicit,elenethk@pes) reper ase eee eee eeee eer er eer eeeeeeeeceetir 023 HirstephalonxyofethindEaicitynwidthofeproximaluendin penser eerie eteitser eee eecrt rer 012 hirdsphalanxsofathirddi citpwlenethe-errecciceridcerl eee ceecr eerie ericerce eer rter O17 Mhirdsphalanxqothirdedi sitmwidtheotsproximalwendeeee see eee eerettsctcice cece recor .008 MERYCHYUS Leidy. Proc. Acad. Nat. Sct. Phila., 1858, p. 24. MERYCHYUS ZYGOMATICUS Cope. (Syn. Ticholeptus zygomaticus Cope, Amer. Naturalist, Vol. XU, p. 129.) The type specimen of this species is a much crushed and distorted skull, of which, through the courtesy of Prof. Cope, I present a drawing (Pl. V, Fig. 45), corrected so far as is possible by the aid of other material. On comparing this skull with that of Mesoreodon, we are at once struck by the great increase in vertical height, both relative and actual, which it has undergone, the height measured verti- cally from the lower border of the mandible to the upper line of the forehead above the orbit being to the total length of the skull about as 7:11, while in Mesoreodon the height barely exceeds one-half of the length. The face has also become some- what shortened and the cranium relatively longer. The orbits do not extend so nearly to the upper line of the skull, the forehead rising much more above them and is more convex, which appears to be due to a greater development of the frontal sinuses. ‘The supraorbital ridges converge less rapidly, and the forehead is thus longer, higher and more arched, and the sagittal crest is shorter. ‘The upper contour of the skull is more arched from before backward, and in all probability the great THE MAMMALIA OF THE DEEP RIVER BEDS. 147 posterior wing-like prolongations of the parietals and supraoccipital are much re- duced. The occipital condyles project much more strongly backward and somewhat more downward. The paroccipital processes are much wider, at least at the base, where they form broad, thin, antero-posteriorly compressed plates. The notch be- tween the postglenoid and posttympanic processes of the squamosal is greatly wid- ened, especially above; below, it is narrowed by the downward and forward course taken by the posttympanic. This change is accompanied by another in the position of the external auditory meatus, which is “directed posteriorly in a way quite pecu- liar, resembling somewhat the position seen in some of the hogs” (Cope). The zygomatic process of the squamosal has a more massive and rugose exter- nal border and its posterior expansion is directed more horizontally and less vertically than in Mesoreodon, and in the latter respect this species departs considerably from the other species of Merychyus. In spite of this difference, there is a very suggestive resemblance between the present species and Mesoreodon in the appearance of the zygoma. The malar is heavier and has greater depth beneath the orbit. The maxil- lary has a shorter but much higher facial portion and the ridge which runs forward from the malar suture is much better marked. In the type specimen, the region about the infraorbital foramen is much injured, but there is some reason to think that the foramen is double, and a facial vacuity is obviously present, and though its shape and size cannot be accurately determined, it was probably small and fissure-like. The premaxillaries are codssified and have their anterior faces much flattened ; the ascending rami are much shorter than in Mesoreodon and the anterior nares lower and more obliquely inclined upward and backward. The muzzle is relatively broader, but the increase in height of the alveolar portion of the premaxillaries and in breadth of their ascending rami, contracts the narial opening, especially in its inferior portion, where it becomes very narrow. The mandible is different from that of Mesoreodon in several important respects. The horizontal ramus is proportionately shorter, but of greater and more uniform vertical depth, tapering less anteriorly ; the chin is straighter and less concave when seen in profile and the incisive alveolus less depressed and procumbent. As in Mes- oreodon, the anterior mental foramen is placed beneath p. , whether, as in that genus, there is a second foramen underneath m. 1, the specimen is too much fissured to show with clearness, though this appears to be the case. The angle projects somewhat below the inferior border of the horizontal ramus and its posterior margin is more thickened and rugose; on the other hand, it projects more behind the condyle, from which it is separated by a more decidedly marked notch. The masseteric fossa is notably smaller and does not descend so low upon the side of the jaw. 148 THE MAMMALIA OF THE DEEP RIVER BEDS. The extreme fragility of the type specimen and the hardness of the cement-like matrix in which it is imbedded have prevented the complete exposure of the teeth, so that these can be studied only from the external side, and important questions as to the constitution of their crowns must be left undecided. In the other specimens of the species at my command, the teeth are so badly preserved as to be of no value in this connection. The first and second upper incisors are very small, the lateral much exceeding the others in size, and the second somewhat larger than the median. These teeth, except for the greater relative size of the lateral one, closely resemble the incisors of the immature skull mentioned above and referred doubtfully to Mes- oreodon intermedius. ‘The canine is of the type usual in the family. The premolars have increased in vertical, but diminished in antero-posterior diameter; their front and hind margins are nearly parallel and the acute apex in each tooth is in advance of the median vertical line of the crown. Thus the postero-inferior cutting border is longer than the antero-inferior. In MJesoreodon the apex is in the middle line and the inferior trenchant margins are of subequal length. The molars exhibit the same reduction in length (antero-posteriorly) and increase in height as do the premolars, and are almost as distinctly hypsodont as are the molars of the later species of Merychyus from the upper Loup Fork. The external pillars, especially the median one (mesostyle), are thin and compressed, but very prominent. The postero-external crescent is much more extended from before backward than the antero-external one, though the disproportion is less than in Mesoreodon. The lower incisors are larger and, in particular, higher than the upper ones, and have compressed chisel-shaped crowns; the canine is broader than the incisors but has lost its typical shape. The caniniform first premolar calls for no remark, as it departs in no way from the shape common throughout the family. The second and third premolars resemble those of the upper jaw in their reduced length and increased height and in having their apices in front of the middle line of the crown. In p.4 the heel (metaconid) is very distinctly separated from the protoconid, but is relatively smaller than in Mesoreodon. The lower molars are so concealed in the matrix that little can be made out with regard to them. As compared with the corresponding teeth of the older genus, they have shorter but much more decidedly hypsodont crowns. MERYCHYUS PARIOGONUS? Cope. Proceedings Amer. Philos. Soc., Vol. XXI, p. 542. In the Princeton collection is the facial portion of a skull, which was found by Mr. Benet in the upper series of beds in the Deep River valley, and which clearly THE MAMMALIA OF THE DEEP RIVER BEDS. 149 belongs to Merychyus. It may be provisionally referred to M. pariogonus Cope, though unfortunately a comparison with the type specimen is of little service, for the two skulls have almost nothing in common. In the type specimen, the cranium and molar teeth are preserved ; in the specimen before us, only the face. The teeth are in such different stages of wear in the two specimens that they may or may not pertain to different species. Almost certainly it cannot be referred to M. zygomaticus, from the type of which it differs in its smaller size and less pronouncedly hypsodont dentition. The incisors, canines and first premolar in the upper jaw are represented only by alveoli, but these show that the canine was rather slender and was followed with hardly an appreciable diastema by p. 1, which was implanted by two distinctly separated fangs, the anterior one of which is considerably larger than the posterior. The second pre- molar has almost plane external face, with no median ridge and only the faintest trace of a cingulum. The crown is so much abraded that only a small enamel invag- ination on the inner side of the tooth remains visible. So far as can be judged from its present worn condition, the construction of p.3 is very much the same as that of p-2, but with a more prominently developed deuterocone. As in the genus, and indeed the family, generally, p.4 consists of two crescents; a difference from the species of the upper Loup Fork is to be found in the strongly developed internal cingulum. The molars increase in size from m. 1 to m. 3; they appear to be as brachyodont as in the type specimen, though this may be due, in part at least, to their abraded condition. The external pillars (para- and mesostyles) are less prominent than in M. zygomaticus. The face has not attained that great vertical depth in the region of the orbits which is characteristic of MZ. zygomaticus ; the orbit is more oval in shape and more oblique in position than in that species and is notched superiorly ; the forehead is more inflated by the sinuses. There are no supraciliary ridges, and the sagittal crest must have commenced at a point considerably farther back than its origin in MMes- oreodon. The malar is not so heavy as in M. zygomaticus, and the masseter ridge is continued well forward upon the maxillary, which displays a slight facial depression above p.2 and p.3. The infraorbital foramen is double; the antero-superior foramen is above p.3 and the postero-inferior opening above the space between p 3 and p.4. The lachrymal is large and has an extensive but not very deep depression. A facial vacuity was obviously present between the maxillary, lachrymal and frontal, but its size and shape cannot be determined, as none of the bones mentioned have complete margins and the nasals are entirely lost. The anterior nares are low, very much nar- 150 THE MAMMALIA OF THE DEEP RIVER BEDS. rowed inferiorly and expanding into a transverse oval above. The premaxillaries have a high and broad alveolar portion which forms an abruptly truncated muzzle ; the two bones are ankylosed at the symphysis. The ascending rami are short and have broad, flattened anterior faces. The palatine processes of the premaxillaries are very limited, while the incisive foramina are quite large and extensively emargin- ate the palatine plates of the maxillaries. The last-named processes are both long and broad and of nearly uniform width, the two series of teeth diverging but little posteriorly. The palatines are very large and form nearly half of the roof of the mouth. ‘Transversely, they extend almost to the molar alveoli, and anteriorly they are carried as far as the middle of m. 1; they are of uniform width, except behind the molars, where they are constricted by the broad palatine notches. The posterior nares are not indicated in the specimen, but obviously they were. placed far back, as the palatines are preserved for some distance behind the last molar. Measurements. M. iene thvofsemolarpremolariscriesi ark ome cic bacriictiereeeree ei ner eecisictiacicierrilets 084 LADEN OF (MEMO GATES 5 cocconsdocsoenb oo copcenaHadccdaocgS ose DaE So DSaD0000 goood odeosoe 041 LORIN, OP DeBaococasacooscacacoscocscccvssv0d es cu pgnc nasa ogo0N0NSoDo OOO SBO ES SoCo OdENSOD -010 Widtitorl Rae A ates LANGE ARM eels, Le eeerise ed th ea anNA eine UR ea .006 Length of p.3.....--.- se ceee eens cece eee ce tenet ee cnet eee rece eee eens O11 \ GWG oi pEsisesseondonanakodonmocHpooUpersaseacD a docdepeEduasd ousuucpe dane anoncauancauate O11 Trengtltrof pe ditaat a yates cae, Onc th ace ne cia ce ear Se Oe age a eR .009 WEI OF 194) so oenccsenq500900cb 00030 9000509 son nos 0gnS 00 SngasoanoddesonesoasScHD0DDbGaG0RE .012 IDEMPAN OF WO WNOEV SABE. on bo co sv oD co Oa DDaB cD DEGSDD DODD SDDNDOODOONGEH DOOD abOOOUEEUSCOSDO 043 DEM RIN OF Ms Wooodoonavcaceod oogatvssusbocusa gobo nessa cuUOD OOO OOb OO ODNDODDODOwOOEOOLADODOE 012 Width of mm. to... 151... eee e ce eee eee ete ee ee ee eee tenes e es 012 ILoMeIN, OF i, Bosco, aaoossoqcoss00s000 ps e0ebanass00bocecosSseaNNSeC arene To rahG ops iedon 017 WGI, Of? Ts (oc osonesuds doo nb sactsd00d 06 cbop oO ODoUO DOGO OOD DOD OObaGACOND OOO UDDSONSHOOOOS 013 ORRIN OF Ml Basscodsvosvcsnssdo0ccacddguDs vaDDD OOO ID GADOdOGODODOOGOOOOOECODNND0000 oo0S .020 Width of m. 8.22.2... eee ec eee vee eee nee et erect eet eres seers: 014 MERYCOCHCIRUS Leidy. Proceeds. Acad. Nat. Sci. Phila., 1858, p. 24. The collection contains nothing from the lower Deep River beds which can be confidently referred to this genus, but there can be no doubt that they occur in those beds because of their abundance both in the John Day and in the Loup Fork. On the other hand, the upper strata of the Deep River valley yielded specimens of Mery- cocherus in profusion. Among these specimens there is great variation in size, and other characters as well, so great as apparently to indicate more than one species, but the only one which can be definitely identified is the JZ. montanus Cope. THE MAMMALIA OF THE DEEP RIVER BEDS. 151 MERYCOCHGRUS MONTANUS Cope. Proceedings Amer. Philos. Soc., Vol. XXI, p. 531. This species is one of the most abundant and characteristic forms of the upper beds and is represented in the collection by specimens which include nearly all parts of the skeleton, the only important structures which are absent being the scapula and pelvis. The skull and dentition of this genus are well known, and for this species in particular have been very fully described by Cope. It will suffice, therefore, for our present purpose to give a brief summary of the more characteristic features of these structures, with especial attention paid to the numerous variations displayed by the various specimens. As in the other members of the genus, the face is bent downward upon the basi- cranial axis and is much more elongate than in the other genera of the family, all of which are characterized by short faces, and in some this shortening is extreme. The brain-case, on the other hand, is relatively short and well rounded. The orbit is small and situated high in the face. The occiput is high and narrow above, though broader than in the other species of the genus in which this region of the skull is known, and becomes very wide at the base. Above the foramen magnum is a nar- row but strong median convexity, which is bounded on each side by a deep fossa and which passes superiorly into the shallow concavity which is enclosed between the wing-like processes of the supraoccipital and parietals.’ The latter processes project much less strongly backward and more transversely than in the older genera of oreo- donts or than in the John Day species of Merycocherus. ‘The occipital condyles do not project very strongly behind the plane of the paroccipital processes, though in this respect there appears to be considerable individual variation. The foramen magnum is unusually high and narrow. The paroccipital processes are very broad at the base, but long, tapering, slender, and antero-posteriorly compressed in the free portion. The tympanic bulle are large, especially from before backward, extending anteriorly beyond the line of the postglenoid processes and nearly to that of the glenoid articular surfaces; between the tympanics the basioccipital is narrow and compressed and has a strong inferior keel. The squamosal forms most of the side wall of the cranium and sends off a massive zygomatic process, which, however, is not so heavy as in the John Day species, M/Z. macrosiegus. The different specimens in the collection exhibit considerable variation in regard to the weight of the zygo- matic process, which does not appear to be of a sexual character, for the differences are not correlated with the size of the canine teeth. In the type of JZ montanus, A. P. S.—VOL. XVIII. T. 152 THE MAMMALIA OF THE DEEP RIVER BEDS. “the zygoma as far as the anterior border of the glenoid cavity is slender, and not convex, but flat in every direction, nor is it decurved, as in WM. superbus. The zygo- matic foramen is relatively much smaller than in that species, but is oblique outwards and forwards at an open angle. The obtuse median edge of the zygoma looks upwards, not outwards, as it does in M. superbus and M. macrostegus, and the supe- rior expansion is opposite the internal extremity of the glenoid face, instead of the external, as in M. superbus, or the middle, as in MZ. macrostegus” (Cope, No. 7, p. 531). None of the specimens in the Princeton collection agree altogether with the type in the construction of this region of the cranium. The description applies best to an almost perfect skull, which, however, differs from the type in the outward and back- ward direction of the posterior or preglenoid boundary of the “zygomatic foramen.” The shape of this opening is different from that seen in the two John Day species which have been mentioned in having its longer axis directed antero-posteriorly, while in them it is transverse. In this specimen the massive and rugose superior expansion of the zygoma is, as in M. macrostegus, above the middle of the glenoid cavity, not above its internal edge, as in the type of the species. A second specimen has the long axis of the zygomatic opening transversely directed, as in the John Day species, and the zygomatic process with its superior expansion is almost as heavy and rugose as in M. macrostegus. This specimen may perhaps represent another species, but the variation in this region of the skull is so great that species can be distinguished only with difficulty. Other specimens show differences of greater or less degree from the two which have been described, which increases the difficulty of distinguishing species. The jugal is flattened and beneath the orbit has great vertical depth. The orbits present more laterally and less obliquely forward than in MZ. macrostegus ; they are also smaller and less prominent than in that species. The forehead is much nar- rower than in the John Day form and made more convex by the enlargement of the frontal sinuses and is decurved at the orbits. In M/Z. macrostegus this decurvature does not occur; the forehead is very wide and almost perfectly flat. The lachrymal is very large and has a deep pit which is very much better marked than in IZ ma- crosiequs. The facial portion of the maxillary is shorter but has a greater vertical height. The nasals are very long and project beyond the alveolar border of the pre- maxillaries, as they apparently do not in WW. macrostegus ; their free ends are obtusely rounded and slightly decurved. They are broader and less decurved at the edges than in the John Day species, but nevertheless are somewhat convex from side to side. The shape of the muzzle is very different in the two species; in the more THE MAMMALIA OF THE DEEP RIVER BEDS. 153 ancient form the premaxillaries are very broad, very much depressed and flattened, and the broad surface presents upward; the anterior nares are rather lower, but the narial notch is deeply incised, extending back over p.2. In Mf. monfanus, on the other hand, the premaxillaries present their broad surfaces anteriorly, not superiorly, which makes the alveolar less depressed and of greater vertical depth; the narial opening is higher, but does not extend so far back, only a little behind the line of the canine. The mandible also differs much in the two species. In M. montanus the horizon- tal ramus is somewhat deeper and has not so straight an inferior border; the sym- physis is longer, straighter and more inclined, less procumbent and projecting at the incisive alveolus. The angle extends to a much less degree and less abruptly below the inferior border of the horizontal ramus, but, on the other hand, it projects farther behind the plane of the condyle, from which it is separated by a more decided notch. The masseteric fossa is smaller, but deeper and more distinctly demarcated. The coronoid is very peculiar; it is short, broad and blunt at the tip, and projects forward much more than upward, in consequence of which the corenoid notch is very broad and shallow. The dentition has no very marked peculiarities. In the upper jaw the incisors are very small, with compressed, simple and obtusely pointed crowns; they increase regularly in size from the first to the third. The canines are very large and of the trihedral shape usual in the family. The anterior premolars have crowns which are low but long antero-posteriorly, with trenchant margins; p.4is less extended in the fore-and-aft direction. The molars increase in size from the first to the third and exhibit little peculiarity of structure. The external buttresses or styles are quite prominent, but the metastyle of m.3, the enlargement of which is so constant a feature in species of Merycocherus as to be of generic value, is smaller than in the John Day species. The para- and mesostyles are prominent, but thin and com- pressed, and the external faces of the para- and metacones are but slightly concave. In the lower jaw the incisors are considerably larger than the corresponding upper series and have high, chisel-shaped crowns; that of the median incisor is very narrow; the second is broader and the third still more so. The canine is functionally one of the incisors, but is much larger than any of those teeth and its crown is pointed rather than chisel-shaped, with obliquely descending supero-external border, which is compressed and trenchant. As in the oreodonts generally, p.1 has assumed the form and function of the canine and is very robust. The crown of p.2 is compressed, trenchant and elongate antero-posteriorly; the deuteroconid is represented by a ridge on the inner face. P. 8 154 THE MAMMALIA OF THE DEEP RIVER BEDS. is similar, except that the inner ridge is more prominent and encloses a shallow pos- terior fossette. The fourth premolar has all the elements of a molar, though owing to the small size of some of these elements it cannot be called exactly molariform. The protoconid is crescentic and there is a small paraconid; the postero-external crescent is formed by the metaconid, which is low and obscurely separated from the protoconid ; the deuteroconid is very clearly demarcated from the protoconid and a low tetartoconid completely encloses the posterior valley or fossette, while the ante- rior valley still opens internally between the para- and deuteroconids. The molars are high and elongate from before backward, and increase in size from m. 1 to m3; the cingulum is developed only on the front and rear faces of the crown, and a tuber- cle between the external crescents represents the mesostylid. The talon of m.3 is large and consists of two crescents separated by a narrow valley. The only specimen of a cranial cast in the collection which can be referred to Merycocherus is not in a satisfactory state of preservation and displays but little that is characteristic. Compared with the brain-cast of Mesoreodon, it exhibits sim- ilar differences to those which are to be observed between larger and smaller species of existing artiodactyls. The hemispheres are broader and of more uniform width, tapering less anteriorly ; the convolutions are but obscurely marked in the specimen and can be interpreted only with difficulty, but the sulci appear to be more sinuous and the accessory sulci somewhat better developed than in Mesoreodon. The anterior portion of the hemispheres is of greater vertical depth proportionately, while the temporo-sphenoidal lobe is relatively rather smaller. The medulla oblongata is large and of subcircular section. The cerebellum is too much damaged for accurate de- scription, but it appears to be rather high and narrow. The atlas is broad, short antero-posteriorly and of robust construction. The anterior cotyli are large both vertically and transversely ; the two surfaces are almost in contact below, but above are separated by a wide triangular notch which emargin- ates the neural arch. The neural canal is relatively small, especially its anterior opening. The sides of the neural arch are steeply inclined and end above in a mas- sive tubercle, which forms a spine of unusual height. The articular surfaces for the centrum of the axis form an angle of about 45° with the median line; in shape they are low and wide and. their medial edges are reflected upon the sides of the neural canal to form the very large continuous facet for the odontoid process. The inferior arch is strongly convex from side to side and displays a small hypapophysial tubercle near the hinder margin. The transverse process extends well forward and has con- verted the atlanteo-diapophysial notch into a foramen. The process is not much ex- tended transversely but widens posteriorly ; its course on the side of the vertebra is THE MAMMALIA OF THE DEEP RIVER BEDS. 155 very obliquely downward and backward, and its antero-external border is thickened and rugose and curved upward in a way that is not found in the other genera of the family ; there is no vertebrarterial canal. The axis is not very well preserved in any of the specimens, but nevertheless some characters of importance may be determined. The centrum is long and de- pressed, with a prominent inferior keel and slightly concave posterior face. The odontoid process is peculiar; for most of its length it is broad and depressed, with flat superior and convex inferior surface and irregularly semicircular free margin ; just before joining the centrum the vertical thickness of the process is suddenly increased, so as to form a step-like elevation on the upper surface along this line. The lateral borders are elevated for a short distance from the centrum, giving the process a spout-shaped section, but at the step already mentioned these raised borders abruptly cease and more than two-thirds of the dorsal face of the process is flat. The neural canal is notably small, with its greatest diameter directed transversely. The neural spine is enlarged into a great plate, which apparently is continued into a posterior spine-like process, as in Hporeodon, though this is not altogether certain. The remaining cervicals have rather short, broad and depressed centra, with opisthoccelous and somewhat oblique articular faces; on the ventral side is a promi- nent keel, which bifurcates behind and into two tubercular ridges, which enclose between them a narrow triangular depression; on each side of the median keel is a deeply concave fossa. The neural canal is very small and of subcircular shape. The neural arch is broad and short and has a nearly flat dorsal surface. The zygapo- physes are broad and flat and present nearly vertically and are but slightly oblique in position; the posterior pair considerably exceed the anterior in transverse breadth. The spine forms a low ridge on the third vertebra, but on the fourth it is well marked and becomes longer on the succeeding vertebre, though it is still very short on the fifth. The transverse processes are variously developed on the different vertebree ; the pleurapophysial plate is very large and massive on the third, and especially so on the fourth, while on the fifth it is smaller and its posterior portion extends outward instead of backward. In all, except the seventh, the vertebrarterial canal is very large. As a whole, the neck of Merycocherus was obviously rather short but heavy, as the structure of the vertebrz indicates the presence of massive and powerful muscles. The thoracic vertebre are represented by several from different parts of the column, belonging to one individual. In the anterior part of the region they have short, broad and depressed centra, with spines much heavier than usually occur in the oreodonts. Posteriorly, the centra become longer and assume the trihedral form 156 THE MAMMALIA OF THE DEEP RIVER BEDS. with slightly opisthoccelous faces which is commonly found among the larger artio- dactyls. In the middle and hinder parts of the region the spines become lower and more slender. The transverse processes in the middle region rise high above the centra and project conspicuously outward. The lumbar vertebra number at least five, since that many are preserved in one specimen. These vertebrz have long, depressed and arched centra, which are quite sharply contracted in the middle. As in the other vertebral regions, the neural canal is notably small, especially in the vertical diameter. The spines are very thin and compressed, but extended antero-posteriorly, and the transverse processes are long and wide but also very thin and depressed. The zygapophyses are of the usual interlocking character, and, in the anterior region at least, the metapophyses are conspicuous. Little is preserved of the sacrum, but enough to show that the pleura- pophyses were massive and apparently confined to the first vertebra. No caudals are represented in any of the specimens: The ribs are known only from a few fragments. It is obvious, however, that the anterior ribs were broader and more flattened than in the other oreodont genera, in which they are remarkably slender for hoofed animals, though probably Merychyus should be excepted from this statement. The humerus is rather short but of massive construction, which is merely an exaggeration of the structure which occurs in the smaller and lighter gencra of the family. The head is large, projecting strongly backward, and is very convex in both directions, so as to be of almost hemispherical shape. The external tuberosity is greatly enlarged and extends across the entire anterior face of the bone, and is strongly curved so as to follow the shape of the head; its free border rises steeply towards the inner side and is wider than the base, projecting at both ends, especially internally, where it is drawn out into a massive hook. The inner tuberosity is small and the bicipital groove very deep. The proximal portion of the shaft is compressed, but of great antero-posterior depth; the deltoid ridge is not very prominent, but extends far down the shaft. The distal end of the humerus is broad and the trochlea is characteristically like that of the other oreodont genera, as is especially seen in the broad, low and rounded intercondylar ridge and in the very prominent and mas- sive internal epicondyle. The ulna is heavy and altogether unreduced, being larger than the radius, except at the distal end. ‘The olecranon is extremely high and deep antero-posteriorly and is thickened and grooved by a tendinal sulcus at the free end. The upper part of the articular surface for the humeral trochlea is broad, but then abruptly contracts and is continued downward only upon the internal side, and the radial facets are distinctly THE MAMMALIA OF TILE DEEP RIVER BEDS. 157 separated by a rather deep groove. The shaft is strongly arched forward and proxi- mally is broad and trihedral; below, it becomes more and more compressed, and in the middle of its course is of almost rectangular section. The distal end is slightly contracted, both transversely and antero-posteriorly, and is almost concealed from view by the radius when seen from the front. The cuneiform facct is simply convex and has its long axis placed nearly in the fore-and-aft direction. The radius retains the family peculiarities in a very marked degree. The head is not greatly expanded and displays the usual three facets for the humeral trochlea. The shaft becomes more and more slender until the middle of its course is reached and then gradually broadens to the distal end. This form of shaft is highly characteristic of the oreo- donts and is in marked contrast to the broad shaft of oval section which is found among the true ruminants. ‘The distal end of the radius is broad, thick and rugose, contracting somewhat suddenly to form the carpal facets; there are no well-marked sulci for the extensor tendons. The surfaces for the scaphoid and lunar are connected in front by a sort of bridge, but for the rest of their extent they are separated by a wide and deep cleft. 'The-seaphoid facet is somewhat oblique in position, is strongly concave from side to side, and is reflected far-up upon the postero-internal angle of the bone. The lunar facet is wider and of a saddle shape, being somewhat convex transversely and concave antero-posteriorly. The carpus of MM. montanus presents some differences from that of the species from the upper Loup Fork, which I have elsewhere described under the name of JV. cenopus (No. 32, p. 346). The scaphoid is large in all its dimensions, but compared with that of the true ruminants, its most striking feature is its great vertical height. The radial surface is rather curiously shaped; the anterior ridge is narrow, but the articular surface descends far down upon the anterior face of the bone and the pos- terior concavity is extended in both directions. The antero-external angle of the proximal end is drawn out into a spur which occupies the “bridge” on the radius mentioned above. The inferior facet for the lunar is very large, both antero-poste- riorly and (near the dorsal side) vertically as well; this facet is but slightly concave. The distal surface is very unequally divided between the facets for the trapezoid and magnum, that for the former being of nearly the same dorso-palmar depth, but much narrower transversely than the latter. There is no articular surface for the trapezium. The lunar is a very curious bone; its radial surface is so warped as to be both convex and concave in both directions ; the anterior border rises steeply towards the ulnar side, where it forms a narrow projection for the cuneiform. The proximal con- tact of the two bones is limited to this small facet, and behind it the upper portion of the lunar is much contracted. On the radial side there is no superior facet for the 158 THE MAMMALIA OF THE DEEP RIVER BEDS. scaphoid, but the inferior one is very large and passes without interruption into the surface for the magnum, which is altogether lateral in position. In MZ. cenopus the magnum facet is strongly convex, but in MW. montanus it is almost flat. The lunar rests entirely upon the unciform and its whole distal surface is occupied by the large concave facet for that bone; it is very oblique in position and forms, with the mag- num surface, a sharp edge or beak which descends almost to the third metacarpal and prevents any anterior contact between the magnum and unciform. The cuneiform has an oblique position in the carpus, running outward and back- ward from the lunar. In shape it is rather low and narrow, but greatly extended in the dorso-palmar direction. The radial side presents two facets for the lunar, of which the proximal one is very small and the distal one quite large. The proximal surface for the ulna is a simple groove and the facet for the pisiform is large, triangu- lar and almost flat. The distal face of the bone is occupied by the large and simply concave surface for the unciform. The trapezium is preserved in Prof. Cope’s type specimen and, as in the case of Mesoreodon, demonstrates that the pollex was not present. It is a small nodular bone which appears not to touch the scaphoid but to articulate merely with the trapezoid and second metacarpal. ‘When seen in position, the trapezoid appears to be as large as the magnum, but is really very much smaller in all its dimensions. The proximal side forms a narrow, imperfectly saddle-shaped facet for the scaphoid, and on the palmar side is a small surface for the attachment of the trapezium ; on the ulnar side the trapezoid is closely applied to the magnum. ‘The metacarpal facets are two in number, a large distal one for the second and a small infero-lateral one for the third metacarpal. The exposed anterior face of the magnum is quite small and yet the bone is a rather large one. ‘The proximal surface is altogether taken up by the facet for the scaphoid, the contact with the lunar being entirely lateral. The scaphoid surface is very convex and rises steeply towards the palmar side to form an ill-defined head. The lunar facet is but slightly concave and in this respect is very different from the surface which occurs in M. cenopus, where the magnum encloses the lunar almost in a semicircle. On the radial side of the magnum is quite a deep concavity which receives the trapezoid. The hook-like process which is given off from the palmar side of the magnum is long and heavy and is strongly recurved towards the radial side. The distal side bears a saddle-shaped facet for the third metacarpal, but there is no anterior contact with the second. The unciform is a large bone of irregularly cuboidal shape. The proximal sur- face is almost equally divided between the facets for the cuneiform and lunar, though THE MAMMALIA OF THE DEEP RIVER BEDS. 159 the latter is somewhat smaller and much more oblique in position, being almost as much lateral as proximal and continuing without interruption into the facet for the third metacarpal, which is relatively large. The distal side bears facets for the fourth and fifth metacarpals, of which the latter is unusually large. The posterior hook of the unciform is long and heavy and projects downward and backward, but is not curved laterally. : In Merycocherus the metapodials attain a degree of shortness and heaviness such as is found in no other genus of the family, and the lateral digits in particular are relatively very stout, so as to recall in some degree the feet of the hippopotamus. The second metacarpal is short and heavy, with a small head, which bears a narrow convex head for the trapezoid and on the postero-external side a small facet for the magnum. ‘There is no anterior contact with the magnum, this bone being excluded from the second metacarpal by the extension of the third to the trapezoid. On the postero-external angle of the head is a small facet for the trapezium. The shaft is somewhat contracted in the middle and is broadest just above the distal trochlea, which is very unsymmetrical in shape. In J. cenopus the second metacarpal is decidedly more slender and compressed than in the species before us. The third metacarpal is much the stoutest bone of the series and its shaft is of almost uniform breadth throughout, though slightly expanding towards the distal end. The head bears a large number of facets and is very completely and perfectly interlocked with the surrounding bones. On the radial side is a flat, horseshoe- shaped surface for the second metacarpal, and above this a small, triangular facet for the trapezoid, which surface, however, is confined to the dorsal moiety of the bone, dying away towards the palmar side and allowing the second metacarpal to reach the magnum. The magnum facet is deeply concave from side to side and convex from before backward; the antero-external angle is drawn-out into a heavy process, which overlaps the head of me. iv and abuts against the unciform. The posterior facet for the fourth metacarpal is large and somewhat oblique, so as to extend slightly under- neath that bone. The distal trochlea is low and broad; its carina is very prominent but does not extend upon the dorsal side. The fourth metacarpal is of about the same length as the third and the shaft has the same antero-posteriorly compressed shape, though it is not so wide transversely. Although the fourth metacarpal is actually no longer than the third, it extends be- yond it distally, for the third rises higher at the proximal end. In Oreodon, me. iii considerably exceeds me. iv in length and projects beyond it both proximally and distally. The surface for the unciform is not very broad and towards the palmar side it becomes obliquely lateral in position. The fifth metacarpal is somewhat A. P. S.—VOL. XVII. U- 160 THE MAMMALIA OF THE DEEP RIVER BEDS. shorter than the second and is heavier, especially towards the distal end; the facet for me. iv is oblique in position. While the median metacarpals are closely applied to and run parallel with each other, the lateral ones, me. ii and me. vy, diverge quite strongly from the median axis, and thus give a foot with a very broad base. The phalanges are short, broad, much depressed and flattened. Unguals are not preserved in any of the specimens. : Except for its increase in size and weight, the femur differs but little from that of the older genera of the family. Compared with the femur of Hporeodon, the fol- lowing changes may be observed. The head is rather more sessile and projects more upward and forward and the pit for the round ligament is larger and deeper. The neck is less constricted and the bridge connecting the head with the great trochanter is thicker and more rugose. The great trochanter, though massive and extended antero-posteriorly, is rather low, not rising so high as the head; it encloses a very deep digital fossa. The second trochanter is smaller and less prominent. The shaft is broader and less rounded and the medullary cavity is larger, with thinner walls. The deep pit above the external condyle for the origin of the plantaris muscle is more conspicuously marked. As in the John Day genus, the condyles are of nearly equal size and are separated by a wide and deep groove. The rotular trochlea is somewhat more modernized, being wider and less symmetrical; the external border is more prominent and continued farther distally than the internal. The proximal end of the tibia is not known.. The distal end is very heavy; the external fossa for the astragalus is considerably broader and less deeply incised than the internal one, and the malleolar process is remarkably long and heavy; on the external side is a small concave facet for the fibula. The fibula has a shaft which is very large in the antero-posterior dimension, though very thin and compressed lat- erally. In the upper Loup Fork species, M/. cenopus, the shaft is much more slender and reduced to almost thread-like proportions. The distal end is expanded into a very heavy external malleolus, which, like the shaft, has its greatest diameter antero- posteriorly. On the inner side is a projection which fits into a groove on the tibia, and distal to this is a large plane surface for the astragalus, which does not, however, occupy all of the tibial side of the malleolus. The calcaneal facet is narrow and slightly concave transversely, but extended in the fore-and-aft direction. The tarsus is lower and broader and the individual elements more massive than in the earlier genera of the Oreodontide, but otherwise there is little change. The astragalus is low and broad; the external proximal condyle is considerably larger than the internal, but the difference is less extreme and the intercondylar groove is narrower and deeper than in MW. cenopus. On the distal trochlea the navicular sur- THE MAMMALIA OF THE DEEP RIVER BEDS. 161 face is wider and the cuboidal narrower than in the last named species. The caleaneum is short and massive and is remarkable for the sessile character of the sustentaculum; in the other members of the Oreodontida: the sustentaculum projects but very little beyond the tuber, though in none of them, except Merychyus, is this so marked as in Merycocherus. The fibular facet is elongate antero-posteriorly, narrow and arched. The navicular is relatively wider than in JZ. cenopus, but is so nearly like that of the older well-known genera as to require no detailed description. The cuboid is low and broad and differs from that of Oreodon in the relative width of the proximal facets, the astragalar being wider than the calcaneal, while in the White River genus the calcaneal is the broader. The difference in MW montanus is not, however, so marked as in the species from the upper Loup Fork. The calcaneal facet is not only narrower than in Oreodon but of different shape, the external border being straight and not projecting beyond the body of the bone. The facet for the astragalus is not so deeply concave as in the more ancient genera of the family, the dorsal and plantar margins not rising so high. On the tibial side are two facets for the navicular, which are separated by a deeper sulcus than in Oreodon. The distal side is almost completely taken up by the large facet for the fourth metatarsal; that for mt. v is very small and rather lateral than distal, while in Oreodon it is altogether distal. The posterior hook of the cuboid is very massive. The entocuneiform is quite a large nodular bone, which articulates with the navicular and mesocuneiform and abuts against the plantar side of the head of the second metatarsal, which it holds firmly in place. As in all the known members of the Oreodontide, the meso- and ectocuneiforms are codssified, and, since the former . has less vertical height than the latter, the compound bone appears to have a step cut in its distal side, which receives the head of the second metatarsal and prevents the third from reaching the mesocuneiform. The metatarsus departs less from the family type than does the metacarpus, both in its proportions and in its mode of articulation with the podials. The median metatarsals are, however, relatively shorter and more massive and the laterals more reduced than in the more ancient genera. The slenderness.of the laterals and their parallelism with the medials are in striking contrast to the lateral metacarpals, which, though shorter than the median pair, are nearly as heavy, and which diverge strongly from the axis of the manus. The second metatarsal is not only proportionately, but even actually, shorter than in the much smaller Oreodon Culbertsoni, and is the shortest of the series. The head articulates with all three of the cuneiforms; on the posterior side is a faucet for the entocuneiform, the proximal end is supported by the mesocu- neiform, and since the latter is of less height than the ectocuneiform, the fibular side 162 THE MAMMALIA OF THE DEEP RIVER BEDS. of the head of the metatarsal is in contact with the ectocuneiform element of the compound bone. The shaft has considerable dorso-plantar diameter, but is very much compressed laterally, and therefore, when seen from the front, appears to be exceedingly slender. The third metatarsal is considerably longer than the second, though relatively very much shorter than in the other genera of the family, and is very massive. The proximal facet is almost plane and articulates only with the ectocuneiform ; a process on the fibular side slightly overlaps mt. iv, but appears not to reach the cuboid; if it does, the contact is very slight. The fourth metatarsal is somewhat longer and rather heavier than the third, and by its broad, plane proximal surface occupies nearly the entire distal side of the cuboid. The fifth metatarsal is somewhat longer and not so compressed and slender as the second; its contact with the cuboid is small and rather lateral than distal. The phalanges do not differ in any important respect from those of the manus ; they are somewhat longer and narrower, and those of the lateral digits are smaller. In particular, the unguals of these digits are very small. CYCLOPIDIUS Cope. Proceedings Amer. Philos. Soc., Vol. XVII, p. 221. The distinction of this genus from Leptauchenia is an obscure one. Cope defines Cyclopidius in brief as being “Leptauchenia without superior incisor teeth;” but this character appears not to be altogether constant, for some specimens show a small alveolus in each premaxillary, and others, described below, have two minute upper incisors on each side. All the peculiarities of Leptauchenia are exaggerated in this genus. The lower incisors are reduced to two in each ramus. ‘The upper canine is usually very small and extends but little below the level of the premolars ; the latter, especially the two anterior ones (p.1 and p.2), are, in the typical species, reduced in size and simplified, but none are lost. The molars are more prismatic than in Lep- tauchenia, and in the upper series the external pillars or styles are more prominent. The first lower premolar retains the form and function of the canine, but is only slightly larger than in the canine proper. The facial region of the skull is much shortened and the vacuities enlarged; the brain-case is small and narrow, but the great expansion of the roots of the zygomatic processes makes the cranial region very broad and low. The auditory meatus is very long and its opening has a more elevated position than in Leptauchenia.. The frontal zone is very short and the frontals form but little of the cranial roof. The nasals are short and slender rods, expanding somewhat anteriorly, where they meet the ascending processes of the maxillaries and premaxillaries; the latter are very small. THE MAMMALIA OF THE DEEP RIVER BEDS. 163 CYCLOPIDIUS INCISIVUS Scott. Amer. Naturalist, 1893, p. 661. This species might with almost equal propriety be referred to Leptauchenia, since it is, in many respects, a connecting link between the two genera; it is much like C. stmus, on the one hand, and ZL. decora of the White River beds on the other. It differs from the latter species principally in the much more reduced incisor teeth and in the larger premolars, and from the former in the presence of two incisors in the premaxillary, the larger upper canine and premolar teeth and in details of skull construction. As in all the species of Leptauchenia, the infraorbital foramen is very small and placed above p.3. The upper incisors, two in number, are extremely small, especially the median one, which is hardly more than a rudiment; the second incisor is almost twice the width of the first and has an obliquely truncate cutting edge. The two incisors of each side are implanted very close together, the first somewhat overlapping the second, while a considerable gap separates the median pair. The canine is larger than in the typical species of the genus, though this character may be sexual, and is followed by a short though distinct diastema, which about equals the fore-and-aft diameter of the canine. The premolars increase in size posteriorly. The first is very small and simple; p.2 has a low internal ridge representing the deuterocone, which in p.3 becomes very distinct and is connected by a ridge with the postero-external angle of the crown, the valley opening in front. The first three premolars have convex external fuces, and are so inserted as to pro- ject slightly backward as well as downward. The molars increase in antero-posterior diameter from the first to the third, m. 3 markedly exceeding m. 2 in this dimension, as the latter exceeds m. 1; in transverse width, however, m. 3 is the least of the series and has much the highest crown. The nasals are more expanded anteriorly at their junction with the maxillaries than even in Leptauchenia decora. The nasal opening is terminal, presenting ante- riorly, and is of heart-like shape, with the apex downward. The premaxille are of somewhat peculiar form; the alveolar portion is insignificant, but the ascending rami form quite a high symphysis and present. their broad surfaces anteriorly, while the superior expansions are twisted, so as to present laterally. The palate is long, broad, and concave from side to side, the two molar series being almost parallel, while the premolars converge anteriorly. The palate is carried farther behind the last molar than in Z. decora and, compared with that species, the posterior nares have been shifted backward. The incisive foramina are very small and anterior in position. 164 THE MAMMALIA OF THE DEEP RIVER BEDS. Measurements. M. Hengthvofimolar-premolar seriesseessceciceeceeee eee ene eee eee ECE ee Eee eee ee eeerer 064 engtheof premolar series rallies ere cesta eee eee eecieamer cence -029 TES ily Wem Sooo os ed oc Goeogaecoc os senosscddococuasaoCd aN sb0Sg0sGKoODONBDC SOS SDESSOCCONS 005 1 AG LD vero itor toRcie eicriaiia ao aortic tia ca boom Sua Goce ead subs bod eo oo ra Sees asead 004 Pits, LENG UD terete teres cistoloitoreiereieqsteystecnexsr csv iieirenfeleinici ne eremsictokeetclsieletet eel nce error cctresierrerec rats 007 Teg 2b WAGNIN poopcoesssoaucodo cc HOS HHGb OOD oso Ee FoDa0 CoDeTODDFOODEOMODEQnOOSEBSOCSONNNN KC 005 Deep be Cttl WeGhmRen ea me nenotatE Canta a on Pond ae mocdonbad oe Genadcoannoeenrawoubaoodsudceus 008 Te ree aaa ee ORES Se eRe AC aay RRP MOE RRS SIS Scheel lta 006 Baten the ha ee doce eee aes te Coon ORIG Ud ne ay ee AE ee a 007 P 4, ACCU eect Sn ree oe Ore Rent ae oh os Sete na ERA Semi A Coe a An UC en Ra Ee ls ea A 007 Tengthwotimolangisericsin ce sseecicrec tence cee Ore ee ner ineccee ere te erate ereccreer 037 Med lem pth ce eeait miss es cisre cictelcierctes ap cvaere nese ote esionerare ete eral omer atsictoeee eer on Sie cher gene eee ee .009 Ree. 7G Kb het oanetaas aos OR RAR noareoTsE 5.5.5 Habis dno Sodnanene anna dig ose Bm cismahabSeldviocs 010 MeEe slengtheecaccrcr kee eee eee reer rreacrer titi ctr sOboabonsoDu odo bacoscbuaysEenhe 013 WG 2, WAGIID o oc Socios suouonD DOD aD SD HaDD So UROSOKOUDODSODO DO LOUHOEGDOUMODGaD BO GGD00000 6800 -010 Wis Ghul Nul Noon Googoscccboduusscucsdodadadupaduoodudauodsudconded cooudbeGcooaddocuKoddedS 015 DM Siawidthiue. oe ohete:, Sahecrnceee Ame tees ee eerie Ai Re 009 The type of this species was found by Mr. R. Stevenson in the upper beds. PitTuHEcistes Cope. Proceedings Amer. Philos. Soc., Vol. XVII, p. 219. In this genus, which is as yet very imperfectly known, the Leptauchenia series of oreodonts appears to have reached its culmination. The lower incisors are reduced to one, the canine has resumed its original functions, and the caniniform premolar has disappeared. 'The other premolars are greatly reduced in size and the mandible is extremely shortened in consequence. Found only in the upper beds of the Deep River valley. Mutuaut RELATIONS OF THE OREODONT GENERA. In my paper upon this family (No. 32), lack of material compelled me to leave many questions with regard to the mutual relations of its genera unsolved and even unattempted. The newly discovered material will.enable us to answer some of these questions with a reasonable degree of probability. We may first consider the origin of Merychyus. The relationship of Mesoreodon to the typical Hporeodons of the Oregon John Day is very obvious and need not be dwelt upon, the only difference of taxonomic value between the two genera being in the structure of the manus, and indeed there is much to be said in favor of giving Mesoreodon only subgeneric rank. Neverthe- less, in the skull and, to a less degree, in the dentition, we may observe numbers of THE MAMMALIA OF THE DEEP RIVER BEDS. 165 slight and subtle changes which are all in the direction of Merychyus. If such spe- cies of the latter genus as MW. zygomaticus and M. pariogonus be taken into account, the transition from Mesoreodon is seen not to be very great or abrupt, though as regards dentition and skull structure there still remains a considerable gap between the two genera, which is only one of the many signs that point to a hiatus between the lower and upper beds of the Deep River deposits. In the Merychyus species from the lower Loup Fork (upper Deep River), M. zygomaticus and M. pariogonus, the face has become deeper and the cranium shorter and the wing-like posterior pro- cesses of the parietals are reduced; the nasals are shortened and a fontanelle is formed between the frontal, lachrymal and maxillary. The premaxillaries are de- pressed, flattened, and ankylosed at the symphysis. In foot structure, Mesoreodon has already attained the condition of Merychyus, especially if the more slender and elongate foot of MZ. intermedius be regarded. In the dentition the principal change consists in a modification of the premolars and a rearrangement of the adjacent horns of the internal crescents on the upper molars, for Merychyus pariogonus shows that the hypsodont molars have been acquired within the limits of the genus. We may, therefore, provisionally at least, regard Mesoreodon as ancestral to Merychyus, and the line of descent would then be: Oreodon—Eporeodon— Mesoreodon— Merychyus. If this view of the case be correct, then the relationship of Merychyus to Mery- cocherus must be strictly one of parallelism, by which the articulation of the third metacarpal with the trapezoid and the depressed and ankylosed premaxillaries have been independently attained in the two genera. Merychyus has also run parallel to Leptauchenia in the development of facial vacuities and in the disposition of the cres- cents of the upper molars as well as their hypsodont character. Yet, now that we know the skull structure of these two genera, no one could seriously maintain that they are genetically connected, though Leidy’s suggestion of such connection was natural enough from the material at his command. To unite Merychyus and Merycocherus into a single genus, as Leidy proposed in his later work (No. 24, p. 201), a sugges- dion which Bettany adopted (No. 1, p. 262), would be to construct an unnatural polyphyletic group, unless genera are to be artificial assemblages united only by certain common characters, the morphological value of which is unimportant. It must be remembered that Merycocherus is a much older form than Merychyus, its peculiarities having all been established in the John Day. 'To derive the latter genus from the former, it would be necessary to make some highly improbable assumptions. (1) We should have to assume that the face had become depressed upon the basi- cranial axis, only to again straighten out and lie in a line with that axis. (2) That the face, after having elongated more than in any other genus of the family, had 166 THE MAMMALIA OF THE DEEP RIVER BEDS. once more become shortened. (3) That the orbit, after retreating backward so as to be almost entirely behind the line of the molar teeth, had again advanced over those teeth. (4) That the zygomatic arches, after having attained an extraordinary degree of size, massiveness and rugosity, had dwindled to proportions even smaller and lighter than those of Oreodon. (5) That the posterior nares had first been pushed back to a remarkable extent, and had again resumed their original position. (6) That the metapodials, after becoming short and massive to a very unusual extent, had attained a degree of length and slenderness which is equally unusual in this family. We have, it is true, already found reason to believe that, in the horses, progress in the main is accompanied by a certain amount of oscillation in the minor details of structure and that even a certain degree of specialization in a direction away from that taken by the phylum as a whole, may be overcome and suppressed, as, for example, in the case of the elbow joint of Mesohippus. Nevertheless, we know of no facts which would justify us in assuming oscillations of such amount as would be involved in the derivation of Merychyus from Merycocherus. If we reject Mesoreodon from the ancestry of the former genus, then we must admit parallelism in the struc- ture of its manus, and thus, whichever horn of the dilemma be accepted, the fact that ‘‘adaptive” reduction of the manus has occurred twice independently within the limits of the family cannot be avoided, for to regard Mesoreodon as in any way descended from Merycocherus is a manifest absurdity. The simplest and most prob- able conclusion is therefore that Merychyus and Merycocherus represent two inde- pendent branches of the oreodont stem, which in some respects have paralleled each other, the former not attaining until the Loup Fork the structures which the latter had already developed in the John Day. Recent discoveries have also thrown some light upon the relationships of the Leptauchenia series. Leidy ascribed that genus to the White River formation, while Cope believed that it was confined to the Deep River beds, though it had not been found in the typical (Montana) locality of that horizon. In my former paper I fol- lowed Cope’s determination, chiefly on the ground that no member of this series has ever been obtained in the John Day beds. It is now proven, however, that Leidy’s determination is the correct one. Dr. Wortman informs me that he has found Leptau- chenia in the upper White River beds, and during the past summer (1893) the Prince- ton party found them in great numbers at the same locality. The morphological difficulty, that all three members of the Leptauchenia series were found only in the same horizon is thus removed, but we are still in the dark with regard to the ancestry of this line, which must be sought for in the lower White River beds. We may, however, confidently remove it from all connection with Merychyus. THE MAMMALIA OF THE DEEP RIVER BEDS. 167 Fam. Indet. HyYPERTRAGULUS CALCARATUS Cope. Bull. U. 8. Geol. and Geogr. Surv., No. 1, 1874, p. 26. A mandible, which is indistinguishable from that of the species named, was found in the lower beds of the Deep River valley, and some specimens from the upper, or Loup Fork, beds of the same locality seem to indicate that the same or a closely allied genus was continued up into the latter series, but the specimens are too fragmentary for certain reference. BLASTOMERYX Cope. U. S. Geogr. Surv. W. of 100th Mer., Vol. IV, PI. II, p. 350. The status of this genus is very’obscure and uncertain. The name was origi- nally applied to m. 3 of a small animal from the upper Loup Fork of Colorado and New Mexico, which appears to be very much like Cosory«, differing from the latter in the shortness of the molar crowns and better development of the basal pillar. So very little is known of the dentition of this animal that its relationships are quite indeterminate beyond the obvious fact of its alliance with Cosoryx. The much larger and more robust species from the lower Loup Fork or Deep River, which has been referred to this genus, not improbably represents a very different one, but materials are lacking for an exact comparison. This Deep River species is in many ways similar to the larger species of Palwomeryx from the upper Miocene of Hurope, and perhaps should be referred to that genus, though in the present state of knowl- edge it would be premature to do so. This doubt is justified by the fact that the mandibular dentition of B. borealis is still unknown, and we cannot therefore deter- mine whether the lower molars possessed the very characteristic “ Palzeomeryx fold,” and it is uncertain whether the type of the Huropean species had developed horns. Schlosser does not regard the presence or absence of horns as a character of generic value, but with this view I am unable to agree. Further, the character of the horns and the shape of the occiput are different from anything which has been observed in the European types. For these reasons, the name Blastomeryx may be provisionally retained. However, by whatever name we call it, there can be little doubt this genus represents a more or less modified migrant from the Old World, not only because of its close similarity, or even identity, with some of the genera of that region, but also because it represents a new element in the American fauna, no form being known from the White River or John Day formations from which it could be derived. That an interchange of mammals between the two continents took place at some time A. P. S.—VOL. XVIII. V. 168 THE MAMMALIA OF THE DEEP RIVER BEDS. subsequent to the John Day and before the beginning of the Loup Fork is made evident by such types as Anchithertum, Blasiomeryx and Mastodon. BLASTOMERYX BOREALIS Cope. Proceedings Amer. Philos. Soc., Vol. XVII, p. 222. This species is quite a large one, though somewhat smaller than the Paleomeryx magnus and P. sansaniensis of Sansan. The skull is remarkable for the high and narrow occiput, the upper portion of which is drawn out into a long, backwardly projecting process composed of the parietals and supraoccipital, which is very similar to the corresponding part of the occiput in the Oreodontide. The horns are trihedral at the base, gradually becoming rounded distally, and are of remarkable length ; they are perfectly simple and unbranched, and in no specimen which I have seen is there any trace of a burr. The surface of the horns is faintly marked by vascular impressions, but is on the whole remarkably smooth, much more so than in the antlers of the deer, and, as Cope has suggested, they were doubtless covered with skin throughout the lifetime of the animal. “At the base of the horn on each side a wing-like expansion extends outward posterior to the orbit ” (Cope). The upper premolars, three in number, have the internal crescent or deuterocone complete ; p.-2 and p.3 are massive and oval in section, while p.4 is more extended transversely. The molars are very brachyodont and are covered with very rugose and strongly wrinkled enamel; the internal crescents are complicated by accessory spurs, which invade the valleys. The internal pillar or style is very variable, being sometimes quite large, while in many specimens it is absent from one or other of the molars. BLASTOMERYX ANTILOPINUS Scott. Amer. Naturalist, 1893, p. 662. The type of this species is represented by a mutilated skull, three cervical ver- tebree and various bones of the fore and hind limbs. It differs from the foregoing species principally in size, being decidedly smaller; the median ribs of the external crescents on the upper molars are less prominent. Other differences are to be ob- served, but they are perhaps rather apparent than real and due to the imperfect con- dition of the specimen. The muzzle is broken away, not only in this, but in all known specimens of B. borealis as well, and hence nothing is known as to the pres- ence or absence of the upper canine. No isolated teeth have as yet been found in the Deep River beds which can be regarded as the upper canines of Blastomeryz, and Filhol reports the same fact with regard to the Palwomeryx of Sansan (No. 13, p- 251). On the other hand, Fraas (No. 16, p. 38) refers the muntjak-like canines THE MAMMALIA OF THE DEEP RIVER BEDS. 169 which he obtained at Steinheim to the species of Palwomeryx occurring at that locality. In the type specimen of B. antilopinus all the premolars are more or less injured ; p.1 appears to be altogether absent; p.2 is represented only by the fangs, but enough remains of p.3 and p.4 to show that they differ only in size from those of B. borealis. Compared with the corresponding teeth of such European species as Paleomeryx mag- nus and P. sansaniensis, these premolars are distinguished by the better development of the deuterocone, the narrower valleys and the character of the cingulum, which is but faintly marked on p.4 and absent from p.3. The latter tooth is the largest of the series, exceeding p.4 not only in antero-posterior but also in transverse diameter. The molars increase progressively in-size from the first to the third, and in all the transverse width is but slightly less than the antero-posterior length, and the cingu- lum is confined to the front face of the antero-internal crescent. The anterior and me- dian external pillars (para- and mesostyles) are prominent. The median rib on the outer face of the antero-external crescent is also conspicuous, though on m. 1 this ridge is less prominent than in B. borealis; the rib of the postero-external crescent is almost obsolete. The internal pillar increases in size from m. 1 to m. 3, being much larger on. m. 8 than on either of the other molars; in B. borealis this pillar is very small or absent on m. 3 and larger on m. 1 and m.2. The internal crescents of the molars are much like those of B. borealis ; the anterior one is less complete than the posterior, its hinder horn being especially shortened; on m. 3 the adjacent horns of the two internal crescents are curiously crenulate, in a way that recalls the transverse crests on the upper molars of some of the extinct horses. In B. borealis this does not appear to be the case. The upper molars of the Sansan species differ from those of the Montana forms principally in the much better developed cingulum, which em- braces the entire crown of the tooth except on its outer side, and in the less devel-' oped internal pillar, which is hardly more than indicated in Filhol’s figures. The inner crescents are not crenulate and are less complicated by spurs which invade the valleys, and the valleys themselves are more widely open. The P. furcatus from Steinheim which Fraas has figured (No. 16, Pl. VIII, Fig. 9) is more like the American species in regard to the structure of the upper molars, so far as can be judged from the drawings. This species is also of interest as showing a mode of formation of the internal crescent of p.2, which I have elsewhere shown to be charac- teristic of Procamelus (No. 34, p. 436), viz, by the coalescence of two distinct ridges in the median transverse line, instead of what is much more usual in the Artiodactyla and universal in the case of p.4, by the extension of ridges from the internal cusp or deuterocone. 170 THE MAMMALIA OF THE DEEP RIVER BEDS. Measurements. M. ILEMEMh Of PENGEVAIMOEIE GOMES cod 00000 boas oo cesses odoocaHacoUdogoUOb ODEN OUCH ODO ODAEOOONNE 079 Wenlathwofepremolarisenicsteee ese ercerreeecretr erence noe rerecerttrs sgoocDDDCdODODODOOS « 037 1a lke) ita den eo eC Mane aa recut an Taare Gea neon ho.anianr Goad GA HACE e Oooo CUE .013 125 8h IGEN oocosspececuce0s 9000000000000 00G000550000000000000000000 000000480000 a5000 000000 013 SPs 45, Meritage, srccaset tess aah cd otatars veleiGial clove otepeeevate sues niche states rasey lace eee toa LoTR otal UE eT ATSC EEN eS 011 lbh OF MME GBT. sagcc obodooccoovcsduosgosonvoDDDaoeS meee iatetatarstelereis/ stone eres 043 Mie Ats ale miethis Serere craigs tratececartierats cierto echoes ieieceae hess Grae Rye Ecler aye Mie pea a aero erage SaaS 014 - Mp wid thanci testers thocecphan testes SAE Coe SAGO AEDST dtd OO MST oMeo oo ose sho ersote pions 018 Nie paket tes litera eres Bat ootmro ane Ono Gone eon mace nee temo SE Donaan oad ooberdeedoodaceusodeS 016 Me WWACUEN oheraye,cjainyeaverss avatars oi cic, oletetatarsieveletatein cdelera ct eteuresstctaie stetaletaralel eater eiaieae che are ear e ere 015 IMR oy al en orth ete tarstotstorelateletelebereretteiierel vate terior ier rereer er ere g500000000006000090000000000 O17 Meds Wi CU airs abarare cote vosareve vere xe aves tecaletcisvel oy sue tolenese ones vloials cco PutegeayeVereuatetapaye eves Screieuete apownte av ave iere pci suarscepetecoye 016 WV. B.—It will be observed that the length of the molar series is less than the sum of the lengths of the individual teeth. This is due to the slight overlapping of the successive molars. The general aspect of the skull (Pl. VI, Fig. 48) is quite similar to that of An- tilocapra, though there are many important differences, which, as would naturally be expected, are in the direction of more primitive conditions. The cranium is very long and the face relatively short, as compared with that of most recent ruminants, though long in proportion to the more ancient forms of the group. In correspond- ence with this, the orbit is placed quite far forward, its anterior rim extending almost to a line above m. 2, and the zygomatic arch is decidedly longer than in the prong- buck. The upper contour of the skull is almost straight, there being hardly any descent at the forehead and little arching of the cranium. In some respects the skull of Blastomerya is more modernized than that of existing hornless deer, such as Hydvopotes and Moschus, especially in the backward shifting of the orbit. In Hy- dropotes the orbit is almost entirely over the molars and in Moschus its front border extends nearly as far as the posterior border of m. 1. In both genera, and especially in Moschus, the orbit is raised much higher above the molar alveolus than in Blasto- meryx. On the other hand, the recent genera have a proportionately shorter and more rounded and capacious cranium, the upper contour of which is much more decidedly arched from before backward, and the occiput is lower and without wing-like exten- sions of the parietals. The paroccipital processes are not advanced in front of the occipital condyles; the zygomatic arch is much shorter and the glenoid cavity more elevated above the plane of the molars, indicating a higher ascending ramus of the mandible. ; The specimens do not indicate that in Blastomeryx the face was bent down the basicranial axis, as in the recent Cavicornia and some other artiodactyls, but appears rather to haye been in the same line with it. The occiput is very broad at the base; THE MAMMALIA OF THE DEEP RIVER BEDS. 171 in the median line, above the foramen magnum, is a wide convexity bounded on each side by a shallow fossa. Towards the summit of the inion this convexity passes into a shallow concavity with a faint median keel, inclosed between processes of the supraoccipitals and parietals. ‘These processes are broken away in the specimen, so that their length cannot be determined. Apparently, however, they were not so long and prominent as in B. borealis, in which the occiput is utterly unlike that of any existing ruminant and has more the peculiar shape characteristic of the Oreodontide. Neither Filhol nor Fraas give figures of this region of the skull in Palewomeryx, but Dicroceros has an entirely different occiput (see Filhol, No. 13, Pl. XXXIV, Fig. 4) which is broad and low and forms a nearly vertical plane. The paroccipital processes in Blastomeryx are long, laterally compressed, and broad at the bases, which are closely applied to the tympanic bulle. Between the condyle and the paroccipital process the inferior surface of the exoccipital displays a large, deep fossa, which is much larger and more deeply impressed than in Anézlo- capra, and the process stands much more in advance of the condyle than in that animal. The mastoid is exposed upon the surface of the skull and forms quite an area between the squamosal and exoccipital; its lower end forms a dense rugose mass, though there is no proper mastoid process. The relations of the mastoid are very much the same as in the prong-buck, except that it 1s more advanced in front of the condyle in position and descends lower upon the paroccipital process. The cranium is long and quite full and rounded, though more slender and less capacious than in Dicroceros. The parietal zone is very long and roofs nearly the entire cranial cavity; obscurely marked temporal ridges pass backward from the bases of the horns and converge to form a low but distinct sagittal crest, which is longer than in the European genus. The postorbital constriction is not strongly marked, though much more so than in the existing genera of horned ruminants. The squamosal is very large and makes up nearly the whole side wall of the cranium; the root of the zygo- matic process forms a thin, depressed plate, which is much extended in the antero- posterior direction and is pierced by a large venous foramen. The zygomatic arch is slender and depressed, and though the distance from the postglenoid process to the last upper molar is nearly the same as in Anfilocapra, yet, owing to the more ante- rior position of the orbit, the zygomatic arch is considerably longer than in the modern genus. The glenoid cavity is thoroughly ruminant in character, though the anterior convexity and posterior concavity are more decided and the postglenoid process somewhat longer than in the prong-buck. The tympanic bull are small and of the shape usual in the antelopes, with a deep groove for the attachment of 172 THE MAMMALIA OF THE DEEP RIVER BEDS. the hyoid apparatus, a feature which is cervine rather than antelopine; the auditory meatus is a long tube which is directed more posteriorly than in Antilocapra. The frontal zone extends considerably in front of the orbits, though but little behind them, and hence takes but a small part in roofing the brain-chamber; these bones lie in nearly the same plane throughout their length, and the descent at the forehead is slight, very much less than in the prong-buck and apparently less than in Dicroceros. Apart from the horns, the upper contour of the skull is thus almost a straight line. The horns are very peculiar and quite unlike those of any other known genus, fossil or recent. At the base the section forms a spherical triangle, the three sides of which present forward, backward and inward; the anterior face is concave, a feature which is much more marked in this species than in B. borealis ; the other faces are convex. In the specimen before us the horns are broken away about three inches above the base, but Prof. Cope’s numerous skulls of the larger species show that in that form, at least, the horns were remarkably long, perfectly simple and non-deciduous, none of them exhibiting any burr or any tendency to branch. The young stages of Dicroceros have a very similar unbranched horn, but the many known skulls of Blastomeryx show that this simplicity is not a transitory character in this genus (see Filhol, No. 13, Pl. XXXIV, Fig. 3). Faintly marked grooves and ridges may be seen on the surface of the horns, but their smoothness indicates, with great probability, that they were permanently covered with skin. The external angle of the base of the horn is in B. borealis continued into a wing-like process which extends outward behind the orbit. In the type of B. antilopinus this process is broken away, but it can hardly have been so prominent as in the larger species. As in Dicroceros and Antilocapra, the horns rise directly above the orbits, but are more erect than in the former genus; the postorbital process is given off from the base of the horn. A large foramen, the supraorbital, pierces the base of the horn and two smaller ones perforate the frontal in advance of the latter; these foramina have a more anterior position than in the prong-buck. Between the frontal and the lachrymal there is a narrow, slit-like fontanelle, the incipient stage of the much larger vacuity which occurs in the deer and many ante- lopes. Cope’s figure of B. borealis (No. 7, Fig. 16) does not show this vacuity ; if it be really absent in that species it will form an important specific distinction. The nasals, premaxillaries, and most of the maxillaries are destroyed, but enough of the latter remains to show that the alveolar portion is very low in correspondence with the extremely brachyodont character of the dentition and that the facial portion is high. In consequence of this, the face is deep vertically, quite as much so as the cranium, and the line from the molars to the occipital condyle is straight and nearly THE MAMMALIA OF THE DEEP RIVER BEDS. alti parallel with the straight upper surface of the skull. There is no trace of a lachry- mal pit in front of the orbit. In B. borealis, and doubtless in the present species also, the maxillaries are sharply constricted in front of the premolars. The infra- orbital foramen occupies a slightly less advanced position than in Moschus, opening above the internal between p.2 and p.3, while in the existing genus it is over the mid- dle of p.2. The palate is broad and gently arched from side to side; between the molars it is of nearly uniform width, but it narrows anteriorly, the two premolar series converging slightly forward. 'The posterior nares extend to about the middle of m. 8 and are very long from before backward, in correspondence with the length of the cranium and zygomatic arches. As contrasted with the base of the skull in Antilocapra, the principal difference to be observed is the elongation of the posterior portion, especially the region between the occipital condyles and paroccipital pro- cesses, which points to a greater development of the cerebellum and medulla oblon- gata and is very usual in the crania of primitive mammals. The orbit is also much lower down and farther forward in the face, its upper border not projecting above the superior contour of the cranium. Nothing is known of the mandible in either species, except some uncharacteristic fragments. From the foregoing description it will be at once evident that while the skull of Blastomeryx is in many respects more primitive than that of any of the recent Pecora, yet it is manifestly of that type and, in some details, such as the character of the occiput and the wing-like processes from the bases of the horns, the genus is special- ized in a way peculiar to itself and which renders it somewhat doubtful whether any existing form is to be derived from it. . Measurements. BLASTOMERYX ANTILOPINUS. ANTILOCAPRA AMERICANA. M. M. Width of occiput at foramen magnum..........-...--..------ 078 079 Distance from crest of inion to middle of horn............... 104 085 Antero-posterior diameter of horn base..............-.--+---- 036 041 Transverse diameter of horn base...........---..-.----------- 046 024 DEIN ORCS Oh My Boacoocassoopg0c00G0dDNOUdsocceoebOOOGKO” 067 -070 Distance from foramen magnum to postglenoid process........ 051 .045 Distance from postglenoid process to m. 3 Bone Borneo 58 5 Coene tC 073 069 Width of palate at m. 3........... 0s erect ee eee eee eee eee 036 052 The vertebral column is represented by the second, third and fourth cervical vertebrae. The axis is completely modernized in character and differs only in details from that of Cervus or Antilocapra. The centrum is broad anteriorly, where it 174 THE MAMMALIA OF THE DEEP RIVER BEDS. expands to form the atlanteal surface; behind this it contracts, to expand again slightly towards the posterior end; the hinder face is concave and there is a strongly marked hypapophysial keel. The articular surface for the atlas does not rise quite so high upon the sides of the neural canal as in Cervus; its inferior border is more curved and the median notch more deeply cut. The neural canal is lower and broader anteriorly, posteriorly its opening is notably small; the pedicels of the neural arch are perforated for the second pair of spinal nerves, but the foramina are smaller than in Cervus ; in its anterior portion the neural arch is thin and plate-like, but gradually thickens until, at the level of the postzygapophyses, it becomes massive and dip- loétic. The spine is so broken that its shape cannot be determined, but it appears to have been thicker and heavier than in Cervus elaphus. The transverse processes are also broken away, but it can be seen that they were slender and probably short. The odontoid process is completely spout-shaped but has a somewhat greater vertical thickness than in the smaller species of Cervus. The postzygapophyses are small and present outward as well as downward. As compared with the axis of Antilo- capra, that of Blastomeryx is of almost the same antero-posterior length, but the surface for articulation with the atlas is wider, the median contraction less pro- nounced and the whole centrum more massive ; the base of the spine is also thicker. But these differences are slight; in general, the axis is very much the same in the two forms. The third and fourth cervical vertebree are likewise of very similar construction to those of the prong-buck; the centra are of almost exactly the same length as in that animal, but the neural arches are somewhat shorter, and thus the gaps between the successive arches are larger; the arches are also distinctly wider transversely. The zygapophyses project more beyond the pedicels of the arch. The neural spines are very low, though better developed than in Antelocapra; on the third vertebra the spine is anteriorly a single ridge, which projects beyond the front of the neural arch and behind bifurcates into two ridges, one running to each of the postzygapophyses. On the fourth the posterior ridges are low, but the anterior rises into a distinct but very short spine. On the corresponding vertebra of the prong-horn these ridges are indicated only in the feeblest way. Measurements. B. ANTILOPINUS. A, AMERICANA. ; M M. Menothwofecentrummotwaxisee eee ereer meer ter iactiaer haa ernest 062 064 \Wiidthiofiantenionita ces tsaxciStee en nere etc eer meen ements .048 044 Mhirdicervicalwlen'= thwofece mt unnepepepetaratete lel letetalatereler tote letetetete reer dL -050 052 Fourth cervical, length of centrum...........- soon ed oOo» suIDDaODD OO COS 054 051 THE MAMMALIA OF THE DEEP RIVER BEDS. 175 The fore limb is represented by a broken humerus, with the ulna, radius and cannon-bone nearly complete. The length of the humerus cannot be determined, as the proximal end is missing, but apparently it was about equal to that of Antzlocapra. On the other hand, the shaft is heavier, especially in the transverse dimension, than in that animal and the deltoid ridge much more roughened and prominent and de- scending farther. The anconeal fossa is deep, but notably small, and the supratroch- lear fossa is shallower and less distinctly marked than in the modern genus. ‘The trochlea is wider and the intercondylar ridge even more prominent and has a more oblique course, downward and inward; the external condyle for the radius is rela- tively somewhat broader; the internal epicondyle is distinctly larger. The distal end of the humerus is much like that of Dicroceros in the less uniform vertical height of the trochlea, which tapers towards the outer side, and in the more external posi- tion of the intercondylar ridge (cf. Filhol, No. 13, Pl. XX XVIII, Fig. 4). The ulna is somewhat more reduced than in Palewomeryx furcatus and less so than in the American antelope, and, so far as can be judged from the only available specimen, was not codssified with the radius at any point. The olecranon is missing, but appears to have projected more decidedly backward than in Antzlocapra. The proximal radio-ulnar articulation is very different from that of the last-named genus, especially in its much greater vertical diameter and in the larger size of the external radial facet, which, however, is set off less distinctly from the body of the bone. The radio-cubital arcade is longer than in the prong-buck, but owing to the shape of the ulnar shaft is narrower. The shaft is very thin and compressed, but proximally has a considerable antero-posterior diameter, which diminishes rapidly as we pass downward. The distal end has but a slight fore-and-aft dimension, but is somewhat thickened transversely and is deeply notched to receive the external angle of the radius. The radius (Pl. VI, Fig. 49) is but little shorter than that of Antelocapra, but has quite a different shape; the lateral and antero-posterior curvatures of the bone are very much as in the recent genus, while the shaft is much broader and less rounded, of oval transverse section and more uniform diameter, much compressed antero-posteriorly, except for the lower one-third of its length. TFilhol’s figure of the radius of Palewomeryx magnus (No. 13, Pl. XXVIII, Fig. 3) shows a very simi- lar shaft, except for a more pronounced lateral flexure which approximates a sigmoid curvature and for a narrower proximal end. The radius of Blastomeryz, so far as its general shape is concerned, is more like that of the fallow deer than of the prong- buck. The trochlea is wider than in the latter, the groove for the intercondylar ridge of the humerus is narrower and emarginates the anterior border more deeply, A. P. §.—VOL. XVIII. W. 176 THE MAMMALIA OF THE DEEP RIVER BEDS. and the ridge external to that notch is wider in correspondence with the more mesial position of the intercondylar ridge. The process for the attachment of the external ligament is a sharp, compressed ridge, which is not so prominent as the massive tubercle of the prong-horn’s radius, and hence the latter, though having a narrower trochlea, measures more across the proximal end than does that of Blastomerya. The distal end differs little from that of the prong-horn, though owing to the broader shaft it expands relatively less; on the anterior face is a broad sulcus for the extensor tendons, bounded by sharp ridges, the inner one of which bifureates near the distal face, forming a second and much narrower sulcus. The carpal facets are very modern in character, except for the less width of the lunar surface, and run very obliquely across the distal face from before backward and mesially; the seaphoid and lunar facets are separated throughout by a sharp ridge and both are reflected far up upon the posterior side of the radius. As in existing ruminants, the radius has expanded so as to come into contact with the cuneiform, though the facet for that bone is much smaller than in the prong-buck. Nothing is known of the carpus, but it may be inferred from the facets of the radius that the lunar is relatively less expanded than in most existing Pecora. _ The metacarpus (PI. VI, Fig. 50) is in the shape of a well-defined cannon-bone, consisting of the coalesced third and fourth metacarpals; no trace of the laterals (ii and v) is preserved, but they were nevertheless probably present in very reduced form, as may be confidently inferred from the condition in Cosoryx. The cannon- bone is considerably shorter than that of the prong-buck, is distinctly stouter and of quite different shape. In Antilocapra, Cosoryx and Blastomeryx gemmifer the proximal end is much compressed in the antero-posterior direction, but in B. antelo- pinus this compression is slight, the transverse diameter but little exceeding the fore-and-aft. The latter diameter diminishes steadily towards the distal end, increas- ing slightly above the phalangeal trochlea; the groove on the posterior face of the shaft is deeper in its proximal portion than in the prong-horn, but is not continued so far down. As in the ruminants generally, the distal venous foramen on the anterior face is extremely small. The trochlez for the phalanges are somewhat lower than in most existing Pecora, but the carine are complete, extending over the entire dorsal face of the trochlee. ; No phalanges are associated with the specimen. Measurements. B. ANTILOPINUS. A, AMERICANA. M. M. uimenussipLreadthsoftadista laendeeeerererenciccceeerreenr re ereeter rnc .041 .038 Humerus, depth of shaft above anconeal fossa...................4e scene 022 017 THE MAMMALIA OF THE DEEP RIVER BEDS. 177 B. ANTILOPINOS. A. AMERICANA. M. M. TRAINS, NEM. 6 scoocoo odo sdddo90d0006500000s0070000 DooagsKg0D0000500000 0 198 205 Radiussabreadthsofsproximalgendsn seer asceeeeeceere reece sirl-ri)- .033 .036 Rachine Iorachin OF Chil GrGlococooccccodouavcdcacconccoddnacdduenc0n00 037 .033 iRaching, loreal oF monllls OF GNA codopnancqbednacnoosGocodso000000D000 024 021 Chynm@maone, WONBNocssgacsobodangedenos0000ndcuososEspsdouEg0q000R0G00 181 206 Cannon-bone, breadth of proximal end............. TRASH MOe GO UA CIOD a6 026 027 Cannon-bonesdepthuofaproximallendesese eee eee ee ereeeeeer reece cree 021 019 @annon=bonesbreadthwotadistalmendsepeeeeeee cee ce cree neiaeeaee cee 032 .026 The femur is badly mutilated, having lost the articular portions of both extremi- ties, and yet the part which remains is longer than the entire femur of Antilocapra ; the shaft is arched forward, compressed and deep, and is decidedly heavier and of less cylindrical shape than in the recent type. The distal portion is trihedral in sec- tion and quite massive; the supracondylar fossa is more deeply marked and rugose and the linea aspera is more prominent. The tibia is very similar to that of the prong-horn and of almost exactly the same length, but heavier and of more massive construction throughout. The spine is lower and less distinctly bifid, the cnemial crest heavier, more prominent and de- scending lower upon the shaft. On the posterior face the roughened lines for mus- cular attachment are much more prominent and rugose. The lower portion of the shaft is less rounded, broader and more oval in section; the distal end is broad and heavy; the fibular facet is altogether distal and shows that the fibula was reduced to a mere nodule. The internal malleolus is very long, the posterior intercondylar ridge or tongue is better developed than in the prong-horn, and the sulcus for the flexor tendons is rather more deeply incised. The tarsus (Pl. VI, Fig. 51) is completely modernized and may be briefly passed over, as it presents but few characters of interest. ‘The astragalus is both higher and wider than in Anfilocapra and in general outline is very similar to Fraas’ figure of Paleomeryx furcatus (No. 16, Pl. VIII, Fig. 13). The pit for the distal median tongue of the tibia is much shallower than in the recent form and the facet for the internal malleolus less deeply incised. The articular surface for the sustentaculum is very large and passes without interruption into the distal trochlea; the latter is almost equally divided between the cuboid and navicular surfaces. The caleaneum is remarkably long, much more so than in the prong-buck or Paleomeryx furcatus ; this elongation, however, chiefly affects the tuber, the portion distal to the sustentaculum being of nearly equal length in all three species. The tuber is deeper (dorso-plantar diameter) and of more uniform depth than in Antilo- 178 THE MAMMALIA OF THE DEEP RIVER BEDS. capra, tapering less towards the free end. The sustentaculum, fibular and calcaneal facets present no noteworthy peculiarity. As in the Pecora generally, the cuboid and navicular are coissified; the com- pound bone is somewhat broader and of about the same vertical height as in the prong-horn. The distal facets on the cuboid portion are quite different from those of Faleomeryx furcatus as figured by Fraas (No. 16, Pl. VIII, Fig. 12); the surface for the main part of mt. iv is broader, especially in front, while that for the posterior hook of the same metatarsal is very much smaller. The length of the hind cannon-bone cannot be determined, as none of the speci- mens are complete. It is evident, however, that it exceeded the fore cannon-bone in this respect more than is usually the case in the prong-horn. The proximal end is of subquadrate shape, the breadth and depth of the head being nearly equal and of the same dimensions as in the specimen of Antilocapra which has been employed for comparison. ‘The hind cannon-bone clearly shows that the portion which articulates with the entocuneiform is the rudiment of the second metatarsal; mt. v is probably represented also, but this is not so obvious. The proximal portion of the shaft is narrow and deep; the groove on the anterior face is strongly marked and terminates distally in a large venous foramen. Measurements. B. ANTILOPINUS. A. AMERICANA. M. M. MemUnp lene theses meee ee amine Eee een (est.) .252 2223 Aili Wen g thier. ters >095060000000000080000000000000050000000000090000000 Tropidoclonvum Cope. aa, No apical papille. f. No preocular plate. Onevinternasall;vanalidividedssscalesikeeledemcyey-tetsicrerlelsieielsleieleleleleletedeletelehelelatelslai-t-Teieieler 2. Fig. 11. Crotalus confluentus Say, Texas. X 2. Plate XV. PEROPODA, ACROCHORDID#, CALAMARINA. Fig. 1. Boa constrictor L., Brazil. X $. Fig. 2. Hunectes murinus L., Brazil. X 3. Fig. 8. Chilobothrus striatus Fisch., Hayti. X 2. Fig. 4. Enygrus bibront D. & B., Fejee Ids. xX 2. Fig. 5. Lichanura trivirgata Cope, Low. California. X 2. Fig. 6. Hryx jaculus L., W. Asia. X 3. Fig. %. Python spilotes Lacep., Australia. x 3. Fig. 8. Ungualia melanura D. & B., Cuba. X 4. Fig. 9. Calamaria gervaisii D. & B., Philippine Ids. x 6. 214 ig. 10. ig. 11. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig g- Fig. Fig. Fig. Fig. Fig, Fig. Fig. for wo = (Ss) so oR oe pO Oe CS eS oprsar we Pp THE CLASSIFICATION OF THE OPHIDIA. Holarchus dolleyanus Cope, Hainan. X 4. Dicraulax purpurascens Schl., Malaysia. xX 4. Pareas moellendorfii Boetch., Hainan. X 3. Both branches split. Acrochordus granulatus Merr., Siam. X 3. Plate XVI. CoLUBRIN &. Drymobdius bifossatus Raddi, Brazil. X 2. Coluber flavescens Laur., Italy. xX 2. Pityophis sayt Schl., W. N. America. X 2. Zamenis ravergiert Menetr., Persia. X 2. Zamenis korros L., Siam. X 2. Cynophis helenw Daud., Ceylon. X 3. Spilotes sebastus Cope, Surinam. X 2. Plate XVII. CoLuUBRIN/®. Compsosoma corais Cuv., Brazil. X 3. Compsosoma virgatum Schl., Asia. X 2. Compsosoma pacilostoma Wied., Brazil. x 3. Gonyosoma oxycephalum Reims, India. X 2. Herpetodryas carinatus Linn., Brazil. X 2. Crossanthera melanotropis Cope, Costa Rica. X 2. Cyclophis modestus Mart., W. Asia. X 4. Contia mitis B. & G., California. ~ 4. Plate X VIII. CoLUBRINZ&. Bascanium flagelliforme Laur., Florida. X 3. Drymobius reticulatus Peters, Peru. X 3. Drymobius boddertii Seetz., Mexico. X 3. Drymobius pulcherrimus Cope, Nicaragua. X 4. Zamenis hippocrepis Linn., Italy. X 2. Entechinus major Gthr., China. X 4. Salvadora bairdit Jan., Mexico. X 4. Macroprotodon cucullatus D. & B, Algiers. x 4. Geagras frontalis Cope, Yucatan. X 3. Ficimia olivacea Gray, Mexico. X 4. Chilomeniscus ephippicus Cope, California. X 4. Stylosoma eatenuatum Brown, Florida. X 3. Hypsiglena ochrorhynchus Cope, Texas. X 4. Plate XIX. CoLUBRINZE. Drymobius margaritiferus Schl., Mexico. X 3. Cacocalyx percarinatus Cope, Costa Rica. X 3. Fig. Fig. Fig. Fig. S Fig. Fig. \ Fig. Fig. to} Fig. Fig. PPS oO HE ge 10. iil, QSOs eo Bs PY SHAS oR wwf Be i ~ aa Re ow THE CLASSIFICATION OF THE OPHIDIA. Cyclophis aestivus L., N. America. xX 4. Phyllorhynchus browntt Stejn., Arizona. x 4. Leptophis ahetulla L., Brazil. x 3. Leptophis prestans Cope, Central America. X 3. Thrasops flavigularis Hallow., W. Africa. X 3. Dendrophis picta L., India. xX 3. Bucephalus capensis Thunb., 8. Africa. x 3. Dasypeltis palmarum Leach, W. Africa. x 4. Cemophora coccinea Blum., Florida. x 4. Plate XX. CoLUBRIN &. Trimetopon pliolepis Cope, Costa Rica. x 4. Conopsis nasus Gthr., Mexico. x 4. Osceola elapsoidea Holbr., Florida. xX 4. Osceola doliata triangulum Boie, New York. x 4 Ophibolus rhombomaculatus Holbr., D. Cal. xX Ophibolus calligaster Say, Kansas. 3. cS) Ophibolus getulus Linn., N. America. X 2. Coronella girundica Daud., Italy. x 3. Proterodon tessellatus Hallow., Japan. X 3 Dianodon rufozonatus, Cantor, China. X 3. Symphimus leucostomus Cope, Mexico. X 3. Rhinochilus lecontei B. & G., Texas. X 3. Plate X XT. CoLUBRIN@, NATRICIN. Herpetodryas melas Cope, Costa Rica. x 2. Drymobius rhombifer Gthr., Ecuador. x 4. Coluber emoryt B. & G., Texas. x 4. Liopeltis vernalis Harl., United States. x 4. Acanthocalyx ventrimaculatus Gray, W. Asia. x 4. Tylanthera florulenta Geoftr., W. Asia. xX 4. Contia episcopa Kenn., Texas. x 4. Ophibolus californie DeBlv., Lower California. x 2. Adelphicus quadrivirgatus Jan., Centr. America. X 3. Ablabes baliodirus Boie, Malaysia. x 4. EHutenia multimaculata Cope, Chihuahua. x 4. Plate X XII. NATRICINE. Owing to the position of the basal section the basal hook was in some cases lost. Natrix rhombifera Hallow., Texas. X 3. Natriz vulgaris Laur., Italy. x 4. Eutenia sirtalis L., N. America. ~x 4. Eutenia melanogaster Wiegm., Mexico. x 4. Natria kirtlandii Kenn., N. America. x 4. Bothrodytes ceylonensis Gthy., Ceylon. x 4. A. P. 8.— VOL. XVII. 2B. 215 16 ig. 10. ig. 11. bad Oo owraa fo we a et od _ a mR wo rt S > 2 Sp Sh he Go StS CO Sb THE CLASSIFICATION OF THE OPHIDIA. Bothrodytes tigrinus Boie, Japan. X 4. Bothrodytes piscator Schneid., India. x 4. Bothrodytes spilogaster Boie, Java. X 4. Storeria dekayt Storer, N. America. x 4. Storerta occipitomaculata Holbr., N. America. X 4. Tropidoclonium lineatum Hallow., Texas. x 4. Plate XXGTT. NATRICINZ! AND HOMALOPSIN. Eutenia proxima Say, Texas. x 4. Natrix septemvittata Say, Pennsylvania. xX 4. Natrix grahamit B. & G., Texas. x 4. Natrix hydrus Pallas, 8. Europe. xX 4. Natriz viperina Merr., Italy. x 4. Natriz stolata L., Hainan, China. x 4. Natrix storerioides Cope, Mexico. x 4. Liodytes allenii Garman, Florida. x 4. Virginia valerie B. & G., Texas. xX 4. Haldea striatula L., Texas. xX 4. Ceratophallus vittatus L., Java. xX 3. Herpeton tentaculatum Lacep., Siam. X 4. Homalopsis buccata L., Siam. X 3. Cerberus rhynchops Schn., India. x 3. Cantoria elapiformis Peters, Siam. X 3. Plate XXIV LyYcopoNnTINz. Lycodon aulicus L., India. x 4. Anoplophallus maculatus Hallow. xX 3. Boaodon virgatus Hallow., West Africa. x 3. Boaodon infernalis Gthr., 8. Africa. x 3. Lamprophis inornatus D. & B., 8. Africa. xX 3. Lycophidium laterale Hallow., W. Africa. x 3. Elapops modestus Gthr., W. Africa. x 4. Dromicodryas bernierti D. & B., Madagascar. x 4. Pseudaspis cana L., South Africa ; the hemipenis in natural erection and not split, one-half not fully evaginated ; from the outside ; @ from above. X 2. Homalosoma luitria L., 8S. Africa. xX 4. Anomalodon madagascariensis D. & B., Madagascar. Plate XXV. DRoMICIN a, Hypsirhynchus ferox Gthr., Hayti. xX 3. Dromicus parvifrons Cope, Hayti. x 4. Ocyophis ater Gosse, Jamaica. X 2. Alsophis angulifer D. & B., Cuba. xX 2. Farancia abacura Holbr., Louisiana. X 2. a oe ie] = dq 39) 8 eI =r >| aS 0s’ dq’ og 0a ey =o ty ue Ftp oe 2 oR iso) ie) lee & os e- a ua oa’ oa oa ie THE CLASSIFICATION OF THE OPHIDIA. Carphophiops amoena Say, N. America. x 4. Fichinanthera eyanopleura Cope, 8. Brazil. x 4. Rhadinea decorata Gthr., Mexico. x 4. Pliocercus elapoides Cope, Mexico. x 4. Ninia atrata Hallow., Mexico. x 4. Tretanorhinus variabilis D. & B., Cuba. x 3. Abastor erythrogrammus Daud., Georgia, N. Amer. x 2, LEPTOGNATHINA. Petalognathus nebulata L., Costa Rica. x 4. Plate XX VI. XENODONTIN. Aporophis anomalus Giinth., Paraguay. X 3. Xenodon almadensis Wagl., Brazil. x 3. Opheomorphus typhlus L., Brazil. 3. a a cobella L., Brazil. x 4. Xenodon severus L., Brazil. x 4. Lystrophis orbignyi D. & B., S. Brazil. x 3. DRoMICIN«. Pseudoeryx plicatilis Linn., Brazil. x 2. Helicops fumigatus Cope, Brazil. x 8. Rhabdosoma badium Boie, Upper Amazon. X 4. as elaps Gthr., Upper Amazon. 3. Acanthophallus colubrinus Giinth., Brazil. x 4. Uromacer oxyrhynchus D. & B., Hayti. x 3. Amastridium veliferum Cope, Panama. x 4. Diadophis regalis B. & G., Arizona. X 4. Plate XX VII. XENODONTINA. Xenodon reginw L. var. Venezuela. X 3. Opheomorphus brachyurus Cope, Brazil. x 4. DRoMICIN A. Hydrops martti Spix., Brazil. x 4. Taeniophallus nicagus Cope, Brazil. x 4. Monobothris chamissonis Wiegm., Peru. X 3. Rhadinea flavilatus Cope, Florida. X 38. ScyTALIN&. Hydrocalamus quinquevittatus D. & B., Mexico. x 3. Philodryas viridissimus L., Brazil. x 4. Thamnodynastes strigatus Gthr., 8. Brazil. x 38. Thamnodynastes nattererti Mik., 8. Brazil. x 4. Tachymenis peruvianus Wiegm., Peru. X 3. Tomodon ocellatus D. & B., Uruguay. xX 4. Rhinostoma nasuum Wagl., 8. Amer. X 3. Fig. Fig. Fig. 14. 15. 16. PAO TP ww p CO DAR NR wD Pp ee = S&S ee ©: ~w oH Co) SY THE CLASSIFICATION OF THE OPHIDIA. Scytale coronatum Schneid., 8. Amer. X 4. Langaha nasuta Shaw, Madagascar. X 3. Grayia smythii Leach, W. Africa. X 2. JEG. XOX TWEE ERYTHROLAMPRINE AND SCYTALINA. Erythrolamprus venustissimus Boie, Brazil. X 3. Conophis lineatus D. & B., Mexico. X 3. Jaltris dorsalis Gthr., Hayti. X 2. Oxyrrhopus plumbeus Wied., Brazil. XxX 3. Oxyrrhopus petalarius L., Brazil. x 3. Philodryas nattererti Steind., Paraguay. x 2. Philodryas schottii Fitz., Paraguay. X 2. Coniophanes fissidens Gthr., Centr. America. X 4. JEU GHG XOKIDR, DRoOMICINE, LEPTOGNATHINE, ERYTHROLAMPRINA AND DIPSADIN®. Heterodon nasicus B. & G., Dakota. X 3. Heterodon platyrhinus Latr., N. America. X 2. Mesopeltis sanniolus Cope, Yucatan. X 4. Leptognathus anthracops Cope, Nicaragua. X 4. Tantilla rubra Cope, Mexico. xX 4. Elapomorphus michoacanensis Dug., Mexico. x 4. Uriechis microlepidotus Gthr., Natal. xX 3. Stenorhina ventralis D. & B., Mexico. X 3. Dipsadomorphus trigonatus Schn., Malacca. 3. Psammodynastes pulverulentus Boie, Tonquin. X 4. Dryophis fulgidus Daud., Centr. America. X 3. Lygophis elegans Tsch., Peru. 2. Letter p, apical disc ; p', same in profile, enlarged. IZ GHG XOXOX, DIPSADIN®. Dipsadomorphus fuscus Gray, Australia. X 3. Dipsas dendrophila Reinwt., Java. XX 2. iTimantodes gemmistratus Cope, W. Mexico. X 4. Rhinobothryum lentiginosum Scop., Brazil. X 3. Sibon septentrionale Kenn., Mexico. xX 4. Sibon nigrofusciatum Gthr., Nicaragua. X 4. Trimorphodon biscutatus D. & B., Mexico. X 2. Crotaphopellis rufescens Gm., Africa. X 2. Chrysopelea ornata Shaw, India. X 2. Procinura emula Cope, Mexico. X 3. Scolecophis atrocinctus D. & B., Centr. America. X 4. Tantilla melanocephala Schi., Brazil. xX 4. Pogonuspis ruficeps Cope, Costa Rica. x 4. Oladophis kirtlandti Hallow., W. Africa. 4. Tragops laetus Cope, Farther India. X 4. Oxybelis acuminata Wied., Centr. America. X 4. THE CLASSIFICATION OF THE OPHIDIA. 219 Plate XXXT. PROTEROGLYPHA AND PLATYCERCA. Fig. 1. Dendraspis jamesonii Traill (Dinophis hammondiit Hallow.), W Africa. x 3. Fig. 2. S&S Crotalus durissus L., Brazil; young. x 3. Nore.—The Anoplophallus maculatus Hallow.,of the preceding pages is the Ophites subcinctus of Boie. There are no hypapophyses on the posterior dorsal vertebrae ; hence the species must be arranged with Lycodon, from which it differs in penial characters. MAY 11 19896 ARTICLE IY. OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. PART II. BY H. V. HILPRECHT, Pu.D., D.D., Professor of Assyrian and Comparative Semitic Philology and Curator of the Babylonian Museum in the University of Pennsylvania. Read before the American Philosophical Society, January 17, 1896. PREFACE. The publication of the history of the American Expedition to Nuffar, announced in the Preface to the first part of the present work, has been delayed by unforeseen circumstances. In view of the increased interest! in these excavations, it seems now necessary to summarize the principal results” and submit them to a wider circle of students. The expedition left America in the summer, 1888, and has continued to the pres- ent day, with but short intervals required for the welfare and temporary rest of the members in the field and for replenishing the exhausted stores of the camp. The results obtained have been extraordinary, and, in the opinion of the undersigned editor, have fully repaid the great amount of time and unselfish devotion, the constant sacri- fice of health and comfort, and the large pecuniary outlay, which up to date has reached the sum of $70,000. Three periods can be distinguished in the history of the exca- vations. 1Cf. especially the official report on the results of the excavations sent by Hon. A. W. Terrell, the United States Minister in Constantinople, to his government in Washington, summer, 1894. ? For details cf. the ‘‘ Bibliography of the Expedition,’’ in Part I, p. 45. To the list there given may be added Peters, ‘‘Some Recent Results of the University of Pennsylvania Excavations at Nippur,”’ in The American Journal of Archeology X, pp. 18-46, 352-368 (with copious extracts from Mr. Haynes’ weekly reports to the Committee in Philadelphia) ; Hilprecht, ‘‘Aus Briefen an C. Bezold,”’ in Zeitschrift fiir Assyriologie VIL, pp. 386-391 ; Assyriaca, Sections I, III-VI. A brief sketch of the history and chief results of the ‘‘ American Excavations in Nuffar”’ will be found in Hilprecht, Recent Research in Bible Lands, pp. 45-63. A. P, S.—vVOL. XVII. 2.¢. 222 OLD BABYLONIAN INSCRIPTIONS First Campaign, 1SS8-18S9.—Staff: John P. Peters, Director; H. V. Hil- precht and R. F. Harper, Assyriologists; J. H. Haynes, Business Manager, Commis- sary and Photographer; P. H. Field, Architect; D. Noorian, Interpreter; Bedry Bey, Commissioner of the Ottoman Government.’ Excavations from February 6 to April 15, 1889, with a maximum force of 200 Arabs. Principal results : Trigonomet- rical survey of the ruins and their surroundings, examination of the whole field by trial trenches, systematic excavations chiefly at III], V, [and X.° Many clay coffins examined and photographed. Objects carried away: Over 2000 cuneiform tablets and fragments (among them three dated in the reign of King Ashuretililani of Assyria), a number of inscribed bricks, terra-cotta brick stamp of Naram-Sin, fragment of a barrel cylinder of Sargon of Assyria, inscribed stone tablet (Pl. 6), several fragments of inscribed vases (among them two of King Lugalzaggisi of Erech), door-socket of Kurigalzu; ce. 25 Hebrew bowls; a large number of stone and terra-cotta vases of various sizes and shapes; terra-cotta images of gods and their ancient moulds; reliefs, figurines and toys in terra-cotta; weapons and utensils in stone and metal; jewelry in gold, silver, copper, bronze and various precious stones; a number of weights, seals and seal cylinders, etc. Second Campaign, 1S89-1890.—Statf: J. P. Peters, Director; J. 1. Haynes, Business Manager, Commissary and Photographer; D. Noorian, Interpreter and Su- perintendent of Workmen; and an Ottoman Commissioner. Excavations from January 14 to May 3, 1890, with a maximum force of 400 Arabs. Principal results : Hxamina- tion of ruins by trial trenches and systematic excavations at III, V and X continued. Row of rooms on the §. E. side of the ziggurrat and shrine of Bur-Sin IT excavated. Ob- jects carried away: About 8000 cuneiform tablets and fragments (most of them dated in the reigns of Cassite kings and of rulers of the second dynasty of Ur); a number of new inscribed bricks; 3 brick stamps in terra-cotta and three door-sockets in diorite of Sargon I; 1 brick stamp of Naram-Sin; 61 inscribed vase fragments of Alusharshid ; 2 vase fragments of Entemena of Shirpurla; 1 inscribed unhewn marble block and several vase fragments of Lugalkigubnidudu; a few vase fragments of Lugalzaggisi ; 2 door-sockets in diorite of Bur-Sin II; over 100 inscribed votive axes, knobs, intag- lios, ete., presented to the temple by Cassite kings; c. 75 Hebrew and other inscribed bowls; 1 enameled clay coffin and many other antiquities similar in character to those excavated during the first campaign but in greater number. 1D. G. Prince, of New York, was the eighth member of the expedition, but during the march across the Syrian desert he fell so seriously sick that he had to be left behind at Bagdad, whence he returned to America. *These numbers refer to the corresponding sections of the ruins, as indicated on the plan published in Part I, Pl. XV. CHIEFLY FROM NIPPUR. 223 Third Campaign, 1893-1896.—Staff: J. H. Haynes, Director, ete.; and an Ot- toman Commissioner; Joseph A. Meyer, Architect and Draughtsman, from June to November, 1894. Excavations from April 11, 1893, to February 15, 1896 (with an in- terruption of two months, April 4 to June 4, 1894), with an average force of 50-60 Arabs. Principal results: Systematic excavations at III, I, If, VI-X, and searching for the original bed and banks of the Shatt-en-Nil. Examination of the lowest strata of the temple, three sections excavated down to the water level ; critical determination of the different layers on the basis of uncovered pavements and platforms; the later additions to the zigeurrat studied, photographed and, whenever necessary, removed ; the preserved portions of Ur-Gur’s ziggurrat uncovered on all four sides; systematic study of the ancient system of Babylonian drainage; the two most ancient arches of Babylonia discovered ; structures built by Naram-Sin and pre-Sargonic buildings and vases unearthed ; c. 400 tombs of various periods and forms excavated and their con- tents saved. Objects carried away: About 21,000 cuneiform tablets and fragments (among them contracts dated in the reign of Dungi and of Darius II and Artaxerxes Mnemon); many bricks of Sargon I and Narim-Sin; the first inscribed brick of Dungi in Nippur; 15 brick stamps of Sargon I, 1 of Naram-Sin; inscribed torso of a statue in diorite (3 of life size, c. 3000 B.C.) and fragments of other statues of the same period; incised votive tablet of Ur-Hnlil; 3 unfinished marble blocks of Lugal- kigub-nidudu and over 500 vase fragments of pre-Sargonic kings and patesis; ¢. 60 in- scribed vase fragments of Alusharshid, 1 of Sargon, 3 of Entemena; 1 door-socket and 1 votive tablet of Ur-Gur; 1 votive tablet of Dungi; a number of inscribed lapis lazuli discs of Cassite kings; fragment of a barrel cylinder of the Assyrian period ; fragments of an Old Babylonian terra-cotta fountain in high relief; water cocks, drain tiles, a collection of representative bricks from all the buildings found in Nippur; ce. 50 clay coffins and burial urns, and many other antiquities of a character similar to those excavated during the first two campaigns but in greater number and variety. With regard to the wealth of its results this Philadelphia expedition takes equal rank with the best sent out from England or France. The systematic and careful manner of laying bare the vast ruins of the temple of Bél and other buildings in Nuffar, with a view to a complete and connected conception of the whole, is equal to that of Layard and Victor Place in Assyria and something without parallel in previous expeditions to Babylonia. Only an exhaustive study and a systematic publication of selected cuneiform texts, which will finally embrace twelve volumes of two to three parts each, can disclose the manifold character of these documents—syllabaries, letters, chronological lists, historical fracments, astronomical and religious texts, building inscriptions, votive tablets, inventories, tax lists, plans of estates, contracts, ete. The 224. OLD BABYLONIAN INSCRIPTIONS results so far obtained have already proved their great. importance in connection with ancient chronology, and the fact that nearly all the periods of Babylonian history are represented by inscriptions from the same ruins will enable us, in these publications, to establish a sure foundation for paleeographic research. Each of the three expeditions which make up this gigantic scientific undertaking has contributed its own peculiar share to the total results obtained. The work of the first, while yielding many inscribed documents, was principally tentative and gave us a clear conception of the grandeur of the work to be done. The second continued in the line of research mapped out by the first, deepened the trenches and gathered a richer harvest in tablets and other inscribed monuments. But the crowning success was reserved for the unselfish devotion and untiring efforts of Haynes, the ideal Baby- lonian explorer. Before he accomplished his memorable task, even such men as were entitled to an independent opinion, and who themselves had exhibited unusual cour- age and energy, had regarded it as practically impossible to excavate continuously in the lower regions of Mesopotamia. On the very same ruins of Nippur, situated in the neighborhood of extensive malarial marshes and “amongst the most wild and ignorant Arabs that can be found in this part of Asia,”' where Layard himself nearly sacrificed his life in excavating several weeks without success, Haynes has spent almost three years continuously, isolated from all civilized men and most of the time without the comfort of a single companion. It was, indeed, no easy task for any European or American to dwell thirty-four months near these insect-breeding and pes- tiferous Affe] swamps, where the temperature in perfect shade rises to the enormous height of 120° Fahrenheit (= c. 39° Réaumur), where the stifling sand-storms from the desert rob the tent of its shadow and parch the human skin with the heat of a furnace, while the ever-present insects bite and sting and buzz through day and night, while cholera is lurking at the threshold of the camp and treacherous Arabs are planning rob- bery and murder—and yet during all these wearisome hours to fulfill the duties of three ordinary men. ‘Truly a splendid victory, achieved at innumerable sacrifices and under a burden of labors enough for a giant, in the full significance of the word, a monumen- lum cere perennius. But I cannot refer to the work and success of the Babylonian Exploration Fund in Philadelphia without saying in sorrow a word of him who laid down his life in the cause of this expedition. Mr. Joseph A. Meyer, a graduate student of the De- partment of Architecture in the Massachusetts Institute of Technology, in Boston, 1Layard, Nineveh and Babylon, p. 565. ? Layard, 0. c., pp. 556-562. ‘ On the whole, I am much inclined to question whether extensive excavations car- ried on at Niffer would produce any very important or interesting results’’ (p. 562). CHIEFLY FROM NIPPUR. 225 had traveled through India, Turkey and other Eastern countries to study the history of architecture to the best advantage. In May, 1894, he met Mr. Haynes in Bagdad and was soon full of enthusiasm and ready to accompany him to the ruins of Nuffar. By his excellent drawings of trenches, buildings and objects he has rendered most valuable service to this expedition. But in December of the same year his weakened frame fell a victim to the autumnal fevers on the border of the marshes, where even before this the Syrian physician of the second campaign and the present writer had absorbed the germs of malignant typhus. In the European cemetery of Bagdad, on the banks of the Tigris, he rests, having fallen a staunch fighter in the cause of science. ven if the sand-storms of the Babylonian plains should eftace his solitary grave, what matters it? His bones rest in classic soil, where the cradle of the race once stood, and the history of Assyriology will not omit his name from its pages. The Old Babylonian cuneiform texts submitted in the following pages have again been copied and prepared by my own hand, in accordance with the principle set forth in the Preface to Part I. The favorable reception which was accorded to the latter by all specialists of HKurope and America has convinced me that the method adopted is the correct one. I take this opportunity to express my great regret that this second part of the first volume could not appear at the early date expected. The fact that two consecutive summers and falls were spent in Constantinople, completing the reor- ganization of the Babylonian Section of the Imperial Museum entrusted to me; that during the same period three more volumes were in the course of preparation, of which one is in print now;' that a large portion of the time left by my duties as professor and curator was to be devoted to the interest of the work in the field; that the first two inscriptions published on Pls. 36-42 required more than ordinary time and labor for their restoration from ¢. 125 exceedingly small fragments; and that, finally, for nearly four months I was deprived of the use of my overtaxed eyes, will, I trust, in some degree explain the reasons for this unavoidable delay. In connection with this statement I regard it my pleasant duty to express my sincere gratitude to George Friebis, M.D., my valued confrére in the American Philosophical Society, for his un- ceasing interest in the preparation of this volume, manifested by the great amount of time and care he devoted to the restoration of my eyesight. The publication of this second part, like that of the first, was made possible by the liberality and support of the American Philosophical Society, in whose TraNnsac- TIONS it appears. ‘To this venerable body as a whole, and to the members of its Pub- lication Committee, and to Secretary Dr. George H. Horn, who facilitated the print- 'Vol. IX, Tablets Dated in the Reigns of Darius II and Artaxerxes Mnemon, prepared in connection with my pupil, Rev. Dr. A. T. Clay, now instructor of Old Testament Theology in Chicago. 226 OLD BABYLONIAN INSCRIPTIONS ing of this work in the most cordial manner, [ return my heartiest thanks and my warm appreciation. No endeavor has been made to arrange Nos. 85-117 chronologically. Although on paleeographic evidence certain periods will be readily recognized in these texts, the cuneiform material of the oldest phase of Babylonian history is still too scanty to allow of a safe and definite discrimination. In order to present the monumental texts from Nippur as completely as possible, the fragment of a large boundary stone now in Ber- lin has found a place in these pages. For permitting its reproduction and for provid- ing me with an excellent cast of the original, Prof. A. Hrman, Director of the Royal Museums, has my warmest thanks. I acknowledge likewise my obligations to Dr, Talcott Williams of Philadelphia and to Rey. Dr. W. Hayes Ward of New York for placing the fragment of a barrel cylinder of Marduk-shabik-z¢rim and the impression of a Babylonian seal cylinder respectively at my disposal. Ifthe text of the latter had been published before, Prof. Sayce would not have drawn his otherwise very natural inference (The Academy, Sept. 7, 1895, p. 189) that the Hyksos god Sutekh belongs to the language and people of the Cassites.' I do not need to offer an apology for in- cluding the large fragment of Naram-Sin’s inscription (No. 120), the only cuneiform tablet found in Palestine (No. 147) and the first document of the time of Marduk- ahé-irba, a member of the Pashe dynasty, in the present series. In view of the great importance which attaches to these monuments, a critical and trustworthy edition of their inscriptions had become a real necessity. The little legend, No. 151, the translation of which is given in the “ Table of Contents,” will prove of exceptional value to metrologists. At the same time I call the attention of Assyriologists to the interesting text published on Pl. 63, which was restored from six fragments found among the contents of as many different boxes of tablets. Nos. 124 and 126, which were copied during the time of the great earthquakes in Constantinople, 1894, belong to the collection designated by me as Coll. Rifat Bey. Together with several hundred other tablets they were presented to the Imperial Otto- man Museum by Rifat Bey, military physician of a garrison stationed in the neigh- 1 Prof. Sayce’s view rests on Mr. Pinches’s hasty transliteration made in connection with a brief visit to America in 1893 and published in Dr. Ward’s Seal Cylinders and Other Oriental Seals (Handbook No. 12 of the Metropolitan Museum of Art in New York), No. 391, where the Cassite god Shugab (= Nergal, cf. Delitzsch, Kossder, p. 25, 1. 12) was transliterated incorrectly by Shu-tah. I called Dr. Ward’s attention to this apparent mistake and gave tbe correct reading in my Assyriaca, p. 93, note. * A boundary stone. The inscription has suffered much from its long exposure to the rain and sun of Babylo- nia. The original, which the proprietor kindly permitted me to publish, is in Constantinople. The stone is so import- ant that it should be purchased by an American or European museum. My complete transliteration and translation of this text and of Nos. 151 and 152 will appear in one of the next numbers of Zeitschrift fur Assyriologie. CHIEFLY FROM NIPPUR. DO borhood of Tello, and were catalozued by the undersigned writer. His Excellency, Dr. Hamdy, Director General, and his accomplished brother, Dr. Halil, Director of the Archeological Museum on the Bosphorus, who in many ways have efficiently pro- moted the work of the American Expedition, and who by their energetic and inte'li- gent efforts have placed the rapidly growing Ottoman Museum on a new, scientific basis, deserve my heartiest thanks for permitting the publication of these texts, and for many other courtesies and personal services rendered during my repeated visits to the Hast. For determining the mineralogical character of the several stones, I am greatly indebted to my colleagues, Profs. Drs. E. Smith and A. P. Brown, of the University of Pennsylvania. The systematic excavations of the last decenniums have revolutionized the study of ancient history and philology, and they have opened to us long-forgotten centuries and millenniums of an eventful past. Hieroglyphics and cuneiform inscriptions were deciphered by human ingenuity, and finally the brilliant reasoning and stupendous assiduity of Jensen in Marburg have forced the “ Hittite” sphinx to surrender her long-guarded secret. He who has taken the pains to read and read again and analyze the results of Jensen’s extraordinary work critically and sine ira et studio, must necessarily arrive at the conclusion as to the general correctness of his system. J am neither a prophet nor the son of a prophet, but I see the day not very far, when the world will wonder—just as we wonder now when we glance back upon the sterile years following Grotefend’s great achievement—that at the close of the nineteenth century years could elapse before Jensen’s discovery and well-founded structure created any deep interest and received that general attention which it deserves. The beautiful marble slab recently found near Malatia’ has offered a welcome opportunity to test the validity of his theory. But the great deszderatwm seems to be more material than is at present at our disposal. Excavations in the mounds of Malatia would doubtless yield it. But what European government, what private citizens, will furnish the necessary funds? May the noble example given by a few liberal gentlemen of Phila- delphia find a loud echo in other parts of the world, and may the work which they themselves have begun and carried on successfully and systematically for several years in Nippur, never lack that hearty support and enthusiasm which characterized its past history. The high-towering temple of Bél is worthy of all the time and labor 1May 23, 1894, together with two other smaller fragments, and now safely deposited in the Imperial Ottoman Mu- seum. With Hamdy Bey’s permission published in Hilprecht, Recent Research in Bible Lands, p. 160. Cf. also Ho- garth in Recueil, XVII, p. 25 f. The inscription cannot be older than 750-700 B.C. The artist took as his motive a hunting scene from the royal palaces of Nineveh. A critical analysis of the well-preserved text will be given by Jen- sen in the next number of Recueil. 228 OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. and money spent in its excavation. Though now in ruins, the vast walls of this most ancient sanctuary of Shumer and Akkad still testify to the lofty aspirations of a by- gone race, and even in their dreary desolation they seem to reécho the ancient hymn once chanted in their shadow : Shadu rabu tuBel Imharsag O great mountain of Bél, Imkharsag, sha réshashu shamami shanna whose summit rivals the heavens, apsi ellim shurshudi ushshishu whose foundations are laid in the bright abysmal sea, ina matati kima rimi ekdu rabsu resting in the lands as a mighty steer, karnashu kima sharitir luShamash shittananbitt whose horns are gleaming like the radiant sun, kima kakkab shamé nab& malic sihati. as the stars of heaven are filled with lustre. (IV &. 27, No. 2, 15-24.) H. V. Hinprecur. FEBRUARY 15, 1896. INTRODUCTION. I, THE LOWEST STRATA OF EKUR. The vast ruins of the temple of Bél are situated on the E. side of the now empty bed of the Shatt-en-Nil, which divided the ancient city of Nippur into two distinct parts. At various times the space occupied by each of the two quarters differed in size considerably from the other. Only during the last centuries before the Christian era, when the temple for the last time had been restored and enlarged on a truly grand scale by a king whose name is still shrouded in mystery, both sides had nearly the same extent. This became evident from an examination of the trial trenches cut in different parts of the present ruins and from a study of the literary documents and other antiquities obtained from their various strata. As long, however, as the temple of Bél existed, the H. quarter of the city played the more important réle in the history of Nippur. Out of the midst of collapsed walls and buried houses, which originally encompassed the sanctuary of Bél on all four sides and formed an integral part of the large temple en- closure, there rises a conical mound to the height of 29 m.* above the plain and 15 m. above the mass of the surrounding débris._ It is called to-day Bint-el-Amir (“daughter of the prince”)* by the Arabs of the neighborhood and covers the ruins of the ancient ziggurratu or stage tower of Nippur, named Jmgarsag? or Sagash® in the cuneiform 1Layard (Mineveh and Babylon, p. 551) and Loftus (Travels and Researches, p. 101) stated this fact clearly. Not- Withstanding their accurate description, on most of our modern maps the site of the city is given inaccurately by being confined to the E. side of the canal. 2 He cannot have lived earlier than c. 500 B.C,, and probably later. 5 Loftus’s estimate of seventy feet (/. c., p. 101) is too low. ‘Layard, l.c., p. 557. Cf. Loftus, /. c., pp. 102Ff. 5««Mountain of heaven,’’ pronounced later Imursag. Cf. Jensen in Schrader’s Keilinschriftliche Bibliothek III, Part 1, p. 22, note 5, and Hommel, Sumerische Lesestiicke, p. 26, No. 306. 6 « High towering’ (on the ending sh cf. Hommel, /. c., p. 141, 2a). Cf. II R. 50, 5-6 a,b. A third name existed but is broken away on this tablet (4a). For Imgarsag cf. alsoTV R&R. 27, No. 2, 15 and 17. A. P. 8.—VOL. XVII. 2 D. 230 OLD BABYLONIAN INSCRIPTIONS inscriptions (ef. Pls. XXIX and XXX). A number of Babylonian kings’ applied themselves to the care of this temple by building new shrines, restoring old walls and repairing the numerous drains and pavements of the large complex, known under the name of Hkur (“mountain house”).” But the three great monarchs who within the last three millenniums before Christ, above all others,’ devoted their time and energy to a systematic restoration and enlargement of the ziggurrat and its surroundings, and who accordingly have left considerable traces of their activity in Nuffar,* are Ashur- banapal (668-626 B.C.),° Kadashman-Turgu (ce. 1250 B.C.)° and Ur-Gur (ce. 2800 L.C.).". The structures of each of these builders have been, one after the other, cleared, measured, photographed and examined in all their details by Mr. Haynes, the intrepid and successful director of the American expedition during the last four years. He is soon expected to communicate the complete results of his work, illustrated by numerous drawings and engravings, in Series B of the present publication. There- fore, referring all Assyriologists to this proposed exhaustive treatise on the history of the excavations, I confine myself to a brief examination of the lowest strata of ancient Ekur, which will enable us to gain a clearer conception of the earliest phase of Baby- lonian history. Whenever it seems essential, Haynes's own words will be quoted from his excellent weekly reports to the Committee in Philadelphia. UR-GUR. At the time of King Ur-Gur the ziggurrat of Nippur stood on the N.-W. edge of an immense platform, which formed the pavement of the entire temple enclosure. It was laid about 2.5 m. above the present level of the plain and had an average thick- ness of 240m. JIn size,‘ color and texture the sun-dried and uninscribed bricks of 1 Among them Dungi (P}. 52, No. 128, cf. his brick legend in Part III of the present work), Ur-Ninib (Pl. 18, No. 10, and Pl. XXIII, No. 65), Bur-Sin I (P1. 11, No. 19), Ishme-Dagan (PI. 9, No. 17, cf. his brick legend in Part III), Bur-Sin If (Pls. 12f., Nos. 20-22), Kurigalzu (Pl. 20, No. 38), Ramm‘ n-shumusur (PI. 28, No. 81), Esarhaddon (cf. Vol. X of the present work and Hilprecht in Z. A., VIII, pp. 390f.). As to the earliest builders cf. below. 7 Cf. Pl. 1, No. 1,8; Pl. 2, No. 2, 10; Pl. 20, No. 38, 7; Pl. 28, No. 81, 8; Pl. 29, No. 82, 8; Pl. 51, No. 121, 8; also Jensen, Kosmologie, pp. 185ff. 8 With the exception of the unknown builder above referred to, who enlarged the base of the early ziggurrat con- siderably and changed its form entirely by adding a peculiar cruciform structure (each arm being 16.48 m. long by 6.16 m. wide) to the centre of its four sides. Each side appeared to have a gigantic wing. *Cf. Part I, p. 5, note, and Néldeke in Hilprecht, Assyriaca, p. 86, ae 1. > Cf. Pl. 29, No. 82, and Hilprecht in Z. A., VIII, pp. 389ff. ®Of. Pl. 24, No. 8, 8. His brick legend will be published in Part IIL. 7Cf. 1 R. 1, No. 8f., and Pls. 51f. of the present work. 823 X15 4 x 7.7 cm., practically the same size as Ur-Gui’s bricks found in the Buwariyya of Warka, Cf. Loftus, c., p. 168, CHIEFLY FROM NIPPUR. 231 this pavement are identical with the mass of crude bricks forming the body of the zigeurrat, while in size and general appearance they closely resemble the burned bricks which bear the name of Ur-Gur. The natural inference would be that Ur-Gur him- self erected this large terrace to serve as a solid foundation for his lofty temple. Yet so long as the inside of the massive ruins has not been thoroughly explored, there remains a slight possibility that the body of the ziggurrat and the pavement existed before Ur-Gur, and that this king only repaired and restored an older building, using in the manufacture of his bricks the mould of his predecessor. On the basis of the present almost convincing evidence, however, I favor the former view and, with Haynes, doubt very much whether before Ur-Gur’s time a ziggurrat existed in ancient Nippur.’ The base of Ur-Gur’s ziggurrat formed a right-angled parallelogram nearly 59 m. long and 39 m. wide. Its two longest sides faced N.-W. and S.-H. respectively,’ and the four corners pointed approximately to the four cardinal points.’ Three of the stages have been traced and exposed (cf. Pl. XXX). It is scarcely possible that formerly other stages existed above.’ The lowest story was c. 64 m. high, while the second (receding a little over 4 m. from the edge of the former) and the third are so 1The ancient name of the temple, Hkuwr, in use even at Sargon’s time, proves nothing against this theory. On the basis of Taylor’s, Loftus’s and his own excavations, Haynes inclines to the view that Ur-Gur was the first builder of ziggurrats in Babylonia. As these two English excavators however did not examine the strata below Ur-Gur’s ter- races, it will be wiser to suspend our judgment for the present, although the absence of a ziggurrat in Tello favors Haynes’s view. In size practically identical with Ur-Gur’s structure in Muqgayyar (ratio of 3:2). Cf. Loftus, 0. c., p. 129. ’ The longest sides of the ziggurrat in Ur fuced N. E. and 8. W. respectively. Cf. Loftus, /. ¢., p. 128. 4«The N. corner is 12° E. of N.’’ (Peters in Zhe American Journal of Archeology, X, p. 18). The Babylonian orientation wasinfluenced by the course of the Euphrates and Tigris, as the Egyptian by the trend of the Nile valley (Hagen in Beitrage zur Assyriologie I, p. 246, note). The Assyrian word for ‘‘ North,” ¢sh(l)tanu, means “ No. I.” From this fact, in connection with the observation that in the Babylonian contract literature, etc., in most cases the upper smaller side (or front) of a field faces N., it follows that the Babyloniansdooked towards N. in determining the four cardinal points, and accordingly could not very well designate ‘‘West”’ by a word which means originally ““back side’ (Delitzsch, Assyrisches Hind worterbuch, p. 44f., and Schrader in Sitzungsberichte der Konigl. Preussisch. Akudemie der Wissenschaften, 1894, p. 1301) like the Hebrews, who faced E. Besides, it is grammatically scarcely correct to derive 7S, a Babylonian loan-word in the Talmud, from a supposed Babylonian aha(u)rru instead of avurru [for this very reason I read the bird mentioned in II R. 37, 12 e. f., not e-har-shu-nw (Delitasch, 1. ¢., p. 45) but a-MUL-shu-nu=NIW Us (cf. Halévy in Reywe Sémitique IIL, p. 91)]. Consequently the only possible reading is am(o)urru, “‘West,’’ as proposed by Delattre, in view of ™@tuA-mu ri and Wu A-mu-ur-ra in the Tell-el-Amarna tablets (cf. also a Babylonian (sic!) village or town A-mu-wr-ri-vki in Meissner, Bevtrage zum Altbubylonischen Privatrecht, No. 42, 1 and 21). Independently a similar result was reached by Hommel in Zettschrift der Deutschen Morgenlandischen Gesellschaft XLIX, p. 524, note 3. 5 No trace of a fourth story could be discovered, and the accumulation of débr’s on the top of Bint-el-Amir is not large enough to warrant the assumption of more than three stages. In Ur Loftus discovered but two distinct stages (1. ¢., p. 128). 232 OLD BABYLONIAN INSCRIPTIONS utterly ruined that the original dimensions can no more be given.’ The whole ziggur- rat appears like an immense altar, in shape and construction resembling a smaller one discovered in a building to the 8.-W. of the temple. As stated above, the body (and faces) of the ziggurrat consist of small, crude bricks,’ with the exception of the §.-E. side of the lowest stage, which had an exter- nal facing of burned bricks of the same size.’ To preserve such a structure for any length of time it was necessary to provide it with ample and substantial drainage. Thanks to the untiring efforts of Haynes, who for the first time examined the ancient Babylonian system of canalisation critically, we learn that the ziggurrat of Nippur had water conduits of baked brick* in the centre of each of the three unprotected sides. They were found in the lower stage and possibly existed also in the upper? ruined portions. On all four sides around the base of the walls was a plaster of bitu- men, 2.75 cm. wide and gradually sloping outward from the ziggurrat towards a gutter, which carried the water away (cf. P]. X XIX, No.74).’ By this very simple arrangement the falling rain was conducted to a safe distance and the unbaked brick foundations were thoroughly protected. Unlike the ziggurrat of Sin in Ur, which had its entrance on the N.-E. side,* the ascent to the different stages in Nippur was at the S.-E. Two walls of burned bricks,” 3.40 m. high, 16.52 m. long and 7 m. distant from each other, ran nearly parallel,” at 1The surface of these stages ‘‘ was covered with a very tenacious plaster of clay mixed with cut straw,’ in order to protect them against storm and rain. ‘‘In places this plaster is still perfect, while in other places several coatings are visible, plainly showing that from time to time the faces of the ziggurrat were replastered’’ (Haynes, Report of Sept. 1, 1894). 2 Cf. above, p. 230, note 8, ‘‘Traces of decayed straw were discovered in these bricks’’ (Haynes, Report of Feb. 9, 1895). , 3In Ur the exterior of the whole lower story was faced by Ur-Gur with baked bricks (Loftus, JU. c., pp. 129f.), while in Warka ‘“‘unlike other Babylonian structures”’ the lower stage of the Buwariyya ‘is without any external acing of kiln-baked brickwork ’’ (Loftus, J. ¢., p. 167). ‘Hach c. 1 m. wide by 3.25 deep. To judge from the height of the ‘‘ buttresses ’’ in Warka, the true meaning of which Loftus failed to recognize, the lowest stage of the Buwariyya had the same height as that of the ziggurrat of Nippur. Cf. Loftus, 7. ¢., p. 169. ; * Cf. Loftus, U. c., p. 129. 6 This plaster rested upon ‘‘a level pavement of two courses of bricks also laid in bitumen, and was 28 cm. thick where it flanked the walls, and 7.7 cm. at its outer edge’ (Haynes, Report of Feb. 10, 1894). ™The projecting casing wall at the base (1.38 m. high) consists of sixteen courses of (stamped) bricks and was built by Kadashman-Turgu around the three unprotected sides of the ziggurrat. In the middle distance of the picture is seen a section of the latest crude brick superstructure (cf. above, p. 230 and note 8) with a tunnel tracing the face of the lowest stage of Ur-Gur’s and Kadashman-Turgu’s ziggurrat. ®Loftus, J. ¢., p. 129. ® Many of which were stamped with Ur-Gur’s well-known legend I R. 1, No. 9. 10 Where they joined the wall of the ziggurrat the distance between them (7 m.) was 1.65 m. greater than at their outer end. CHIEFLY FROM NIPPUR. 233 right angles from the face of the ziggurrat, into the large open court, which extended to the great fortification of the temple. This causeway‘ was filled up with crude bricks of the same size and mould and formed a kind of elevated platform, from which apparently steps, no longer in existence, led up to the top of the ziggurrat and down into the open court in front of it. The whole temple enclosure was surrounded by a large inner and outer wall built of sun-dried bricks. To the N.-W. of Ekur “30 courses of these bricks are still plainly visible.”* They compose the ridge of the outer wall and, like the pavement of Ur-Gur’s ziggurrat, rest on an older foundation. The complete excavation of the inner wall will be undertaken in connection with the systematic examination and removal of the ruins around the ziggurrat. SARGON AND NARAMGSIN. Immediately below “the crude brick platform of Ur-Gur,” under the H. corner of the ziggurrat, was another pavement consisting of two courses of burned bricks of uniform size and mould.’ Each brick measures c. 50 em. in square and is 8 em. thick. This enormous size is quite unique among the more than twenty-five different forms of bricks used in ancient Nippur, and enables us to determine the approximate date of other structures built of similar material in other parts of the city. Fortunately most bricks of this pavement are stamped. A number of them contain the well- known inscription of Shargani-shar-ali, while the rest bears the briefer legend of Naram-Sin (Part I, Pls. 3 and II). This fact is significant. As both kings used the same peculiar bricks, which were never employed again in the buildings of Nip- pur, and as they are found near together and intermingled in both courses of the same pavement, the two men must necessarily be closely associated with each other. ‘lhis ancient brick pavement becomes therefore a new and important link in the chain of my arguments in favor of the identity of Shargani-shar-Alit with Sargon JI, father of 1 Both the walls of the causeway and those of the ziggurrat were battered, the batter of the former (1:8) being exactly half the batter of the latter (1:4), according to Haynes’s Report of Feb. 9, 1895. Cf. Loftus, J. ¢., p. 128. 3 Haynes, Report of Sept. 8, 1894. 3 Niebuhr’s very recent remarks on the historicity of Sargon I and Naram-Sin (Chronologie der Geschichte Israels, AHgyptens, Babyloniens und Assyriens, Leipzig, 1896, p. 75) should never have been made after the publication of their inscriptions in the first part of the present work. His insinuations against the priests of Nippur read like a carnival joke, in the light of the facts presented in the following sketch. 4Oppert’s proposed reading of this name as Bingant sar-tris (Revue d’ Assyriologie Iil, pp. 25f.) is impossible and was declined in Assyriaca, p. 30, note 1. The original picture of the sign Shar in our name is not ‘‘1’hiéroglyphe de Varbre en feuilles’”’ (Oppert, 7. c.), but an enclosed piece of land covered with plants, in other words a plantation, garden, orchard (kiru). Cf. Bertin, Origin and Development of the Cuneiform Syllabary, p. 7. 234 OLD BABYLONIAN INSCRIPTIONS Naram-Sin' (Part I, pp. 16-19). It was apparently laid by Sargon and relaid by his son, Naram-Sin, who utilized part of his father’s bricks, and it must therefore be rec- ognized as the true level of the Sargon dynasty in the lower strata of the temple at Nuffar. No bricks of either of the two kings have been found below it, nor in fact any other inscribed objects that can be referred to them.” But another, even more powerful witness of Naram-Sin’s activity in Nippur”® has arisen from some ruins in the neighborhood of Ekur. On the plan of Nuffar published in Part I, Pl. XV, a ridge of low insignificant- looking mounds to the N.-W. of the temple* is marked VII. They represent a portion of Nimit-Marduk, the outer wall of the city. Its upper part, as stated above, was constructed by Ur-Gur. During the summer of 1895 Mr. Haynes excavated the lower part of this rampart. He selected a piece of 10 m. in length and soon after- wards reported the following surprising results. The foundation of the wall was placed on solid clay c. 3m. below the water level or c. 5 m. below the plain of the desert. It was “built of worked clay mixed with cut straw and laid up én masse with roughly sloping or battered sides ” to a total height of c. 5.5 m. Upon the top of this large base, which is e. 13.75 m. wide, a wall of the same enormous width, made of sun-dried 1More recently (Altorientalische Forschungen ILI, p. 238) Winckler refers to Shargani-shar-ali as the possible his- torical basis of ‘‘the mythical Sargon of Agade.’’ I trust the day is not very far when he will regard Sargon as histori- cal and identical with Shargani-shar-ali, as I do. 2 The brick stamp of Sargon, mentioned below, p. 243, as having been unearthed underneath the wall of Ur-Gur’s archive, indicates that this underground archive or cellar existed at Sargon’s time at that very spot and was rebuilt by Ur-Gur. ‘ 8Inscribed burned bricks of Naréim-Sin were also found in mound X, on the W. bank of the Shatt-en-Nil at a very low level. All the stamped bricks of Naraém-Sin ‘‘show evident traces of red coloring on their under or inscribed face’’ (Haynes, Report of Nov. 24, 1894). * Originally these mounds continued a little farther N. W.than they can be traced on the map, until suddenly they turned to the W., reaching the Shatt en-Nil apparently not far from II. A large open space, ‘(414 m. long by 276 m. wide and covering more than 26 acres of ground,’’ was enclosed by this wall, by the mounds called VIIL and by the temple complex (III). As far as the present evidence goes, this court was never occupied by any brick build- ings. Its real purpose can therefore only be surmised. According to Haynes (Report of August 3, 1895) it served as a caravanserai for the accommodation and safety of pilgrims and their animals. Such a view is possible, but it seems to me more probable to regard this enclosed place as a court where the numerous cattle, sheep, etc., received by the temple administration as regular income and for special sacrifices, were kept and sheltered. Perhaps it served both purposes. Besides in the time of war the inhabitants of Nippur readily found a safe refuge behind its walls. On the N. E. side of this court, ‘“‘at the foot of the enclosing wall, a bubbling spring was discovered. On either side of the spring are still seen the brick platforms and curbs where the water pots rested.’’ From the size of the bricks, which “appear to be the half bricks of Naram-Sin,”’ the spring existed at the time of this great builder. ‘‘ After the court had become filled to a depth of about 1 m., adiagonal wall of burned bricks, 54 m. long, six courses high, placed on a raised base of clay, was built before the spring to divert the course of drifting sand and débris from the court.’’ 5 Cf. IL R. 50, 29a, b. The inner fortification (d&rw) was called Imgur-Marduk (ibidem, 28a, b). Cf. Delitzsch, Wo lag das Paradies? p. 221. Both names seem to be of comparatively late date and cannot be applied to Naram-Sin’s fortifications. According to IL 2. 50, 30f, a, b, two other names existed for the outer wall (shalhw). CHIEFLY FROM NIPPUR. 235 bricks, was raised to an unknown height.!_ We may well ask in amazement, Who was the builder of this gigantic wall, constructed, as it seems, ana im site? Nobody else than the great Narim-Sin, whom Niebuhr of Berlin finds hard to regard as a histori- cal person! Perhaps this scholar will now release me from presenting “ wirkliche Inschriften politischer und als solcher glaubhafter® Natur, damit man ihrer [namely, Sargon’s and Narim-Sin’s| einstmaligen Hxistenz vollkommen traue.”* The bricks had exactly the same abnormal size as the burned bricks of the pavement below the zigeurrat and, in addition, although unbaked, bore Naram-Sin’s usual stamped inscrip- tion of three lines. “They are dark gray in color, firm in texture and of regular form. In quality they are unsurpassed by the work of any later king, constituting by far the most solid and tenacious mass of unbaked brick that we haye ever attempted to cut our way through.”* To illustrate the amount of time, patience and labor needed for the systematic exploration of these lowest strata, it may be mentioned that one of the sections excavated contained ‘‘more than 60,000 cubic feet ’’ of earth, which had to be carried away in basketfuls a distance of 120 m. and at the same time to be raised to a height of 15-24 m. Haynes, Report of Oct. 5, 1895. A, P. S.—VOL. XVIH. 2B. 238 OLD BABYLONIAN INSCRIPTIONS time. But as no evidence of an ancient ziggurrat previous to Ur-Gur and Naram- Sin has been discovered, the accumulations must have necessarily been slower and yresuppose a longer period than elapsed between Naram-Sin and the final destruction I PI ger I of Ekur in the first post-Christian millennium. I do not hesitate, therefore, to date the founding of the temple of Bél and the first settlements in Nippur somewhere between 6000 and 7000 B.C., possibly even earlier. I cannot do better than repeat Haynes’ own words, written out of the depth of this most ancient sanctuary of the world so far known: “ We must cease to apply the adjective earliest to the time of Sargon or to any age or epoch within 1000 years of his advaneed civilization.”* “The . golden age of Babylonian history seems to include the reign of Sargon and of Ur- Gur 993 Somewhat below the pavement of Nar&im-Sin, between the entrance to the zig- gurrat and the E. corner, stood an altar of sun-dried brick, facing $.-H. and 4 m. long by 2.46 m. wide. The upper surface of this altar* was surrounded by a rim of bitu- men (18 em. high), and was covered with a layer of white ashes (6.5 em. thick), doubtless the remnant of burned sacrifices. To the $.-W. of it Haynes discovered a kind of bin built of crude brick and likewise filled with (black and white) ashes to the depth of c. 50 cm. Ata distance of nearly 2 m. from the altar (in front of it) and ce. 1.25 m. below the top was a low wall of bricks, whose limits have not yet been found. Apparently it marked a sacred enclosure around the altar, for it extended far under the pavement of Naram-Sin ° and reappeared under the W. corner of the ziggur- rat.’ The bricks of which this curb was built are plano-convex in form.* They are laid in mud seyen courses (= 45 em.) high,’ the convex surface, which is “ curiously creased lengthwise,” being placed upward in the wall. At a distance of 4.62 m. outside of this low enclosure and ec. 36 em. below its bottom stood a large open vase in terra-cotta with rope pattern” (cf. Pl. XX VII, No. 72). It will serve as an excellent specimen of early Babylonian pottery in the fifth millennium before Christ. Undisturbed by the hands of later builders, it had remained 1A similar conclusion was reached by Peters in The American Journal of Archeology X, pp. 45f. * Report of August 30, 1895. > Report of August 8, 1895. *Which was 0.92 m. below the level of Narim-Sin’s pavement. > Haynes, Repoit of Feb. 17, 1894 (also Aug. 24, 1895). Haynes’s chemical analysis of the white ashes showed evident traces of bones. ®The facts concerning this curb have been gathered from Haynes’s Reports of Feb. 17 and March 17, 1894; Aug. 3, 1895. 7Cf. Peters, The American Journal of Archeology X, pp. 31 and 44. ® With an average length and breadth of 24.5 x 18 cm. * «Being placed lengthwise and crosswise in alternate courses” (Haynes, Report of March 17, 1894). 10 Haynes, Report of Aug. 24, 1895. CHIEFLY FROM NIPPUR. 239 in its original upright position for more than 6000 years, and it was buried under a mass of earth and d2bris long before Sargon I was born and Naram-Sin fortified the temple of Nippur.* A second vase of similar size but different pattern” was discovered 77 cm. below the former and nearly double the distance from the ancient brick curb. There is little doubt in my mind that both vases, which stood in front of the altar, on its S.-S.-E. side, one behind the other as one approached it, served some common purpose in con- nection with the temple service at the pre-Sargonic time. Another section of earth adjoining the excavation which had yielded these remarkable results was removed by Haynes. To the S.-H. of the altar described above, almost exactly under the E. corner of Ur-Gur’s ziggurrat and immediately below the pavement of Naram-Sin, stood another interesting structure.’ It is 338 m. high,'7 m. square, “with a symmetrical and double reéntrant angle at its northern corner and built up solidly like a tower.” Its splendid walls, which exhibit no trace of a door or opening of any kind, are made of large unbaked bricks of tenacious clay” somewhat smaller in size than those of Naram-Sin’s rampart. While examining the surroundings of this building, Haynes found ten basketfuls of archaic water vents and fragments thereof on its S.-H. side and on a level with its foundation. His curiosity was aroused at once, and after a brief search underneath the spot where the greatest number of these terra-cotta vents and cocks had been gathered, he came upon a drain which extended obliquely under the entire breadth of this edifice. At its outer or discharging orifice he found the most ancient keystone arch yet known in the history of architecture. The question once asked by Perrot and Chipiez® and answered by them with a “probably not,” has been definitely decided by the American expedition in favor of ancient Chaldzea. The bottom of this valuable witness of pre-Sargonic civilization’ was c. 7 m. below the level of Ur-Gur’s crude brick platform, 4.57 m. below the pavement of Naram-Sin, and 1.25 m. below the foundations of the aforesaid building. The arch is 71 em. high, elliptical in form, and has a span of 51 cm. and a rise of 88cm. Cf Pl. XXVIII, 1Tt stood 3.05 m. below the pavement of Naram-Sin. *Tn the form of a large jar, its diameter in the centre being larger than that at the top (Haynes, Report of Aug. 24, 1895). 3 The following facts have been gathered from Haynes’s Reports of Oct. 13, Nov. 24, 1894. 4Tts foundations are therefore 3.38 m. below the level of Naram-Sin’s pavement. 5««Thoroughly mixed with finely cut straw and well kneaded.” 6A History of Art in Chaldea and Assyria, Vol. II, p. 234. 7faynes, Reports of Oct. 13, 20, Nov. 24, 1894; Jan. 12, March 2, 1895. 240 OLD BABYLONIAN INSCRIPTIONS No. 73.’ The bricks of which it is constructed are well baked, plano-convex in shape, and laid in clay mortar, the convex side being turned upward. A few months after its discovery the arch was forced out of shape, “probably from the unequal pressure of the settling mass above it, which had been drenched with rain water.” Whether the altar, the two large vases and the massive building, under which the ancient arch was found, had any original connection with each other, is at present impossible to prove. According to my calculations and our latest news from the field of excavation, the bottom of the lower vase and the foundation of the massive build- ing were not on the same level. The difference between them is nearly 0.5m. As the highest vase, however, stood 77 cm. above the other, and as the section S.-H. from them has not yet been excavated, it is highly probable that a third vase stood at some distance below the second. However this may be,so much we can infer from the facts obtained even now, that an inclined passage from the plain led alongside the two vases to the elevated enclosure around the solitary altar. I am therefore disposed to assign to the tower-like building, the character of which is still shrouded in mys- tery, the same age as the altar, curb and vases. The keystone arch and drain, on the other hand, are doubtless of a higher antiquity. Whether the 5200 years given by Nabonidos as the period which elapsed between his own government and that of Sargon I, be correct or not, the arch cannot be placed lower than 4000 B.C., and in all probability it is a good deal older. The two sections which contained all the buildings and objects described above were carried down to the virgin soil, where water stopped our progress. A third section removed in their neighborhood yielded similar results. But it is impossible to enumerate in detail all the antiquities which were uncovered below the S.-H. side of the ziggurrat. The lowest strata did not furnish any treasures similar te those found in the upper layers; they showed a large proportion of black ashes and fine charcoal mingled with earth, but they also produced many smaller objects of great interest and value, especially fragments of copper, bronze and terra-cotta vessels. Several pieces of baked clay steles, bearing human figures in relief upon their surface, will be treated at another place and time.” An abundance of fragments of red and black lacquered 1A kind of pointed arch of unbaked brick (60 cm. high and 48 cm. wide at the bottom) was found by Haynes in mound X (cf. Pl. XV), on the 8. W. side of the canal bed. From the depth in which it was discovered, Haynes reasoned correctly that it was older than 2000 B.C. From the inscribed objects excavated in connection with it, I determined that it must have existed at the time of the dynasty of Isin (c. 2500 B.C.). In all probability it dates back to Ur-Gur’s period. For the wall in which this arch is placed was built of the same sun-dried bricks which compose the body of the ziggurrat (Haynes, Reports of April 27, Dec. 21, 1895). Tor the general form of this pointed arch cf. Perrot and Chipiez, J. ¢., p. 229, Fig. 92. *One of them was found at a depth of 7 m. below the pavement of Narim-Sin and 2.44 m. lower than the bottom of the aich, within about 2m. of the lowest trace of civilization (Haynes, Report of Sept. 7, 1895). Another was discovered 7.70 m. below Naram-Sin’s pavement (Report of Sept. 14, 1895). CHIEFLY FROM NIPPUR. 241 pottery was discovered at a depth of 4.6 m. to 8 m. below the pavement of Nardm- 1 Sin.t “Had these pieces been found in the higher strata, one would unhesitatingly declare them of Greek origin, or at least ascribe them to the influence of Greek art.” For they are, as a rule, of great excellence and in quality far superior to those found in the strata subsequent to the period of Ur-Gur. The results of our excavations in the deepest strata of Ekur will change the cur- rent theory on the origin and antiquity of the arch, will clear our views on the devel- opment of pottery in Babylonia, and will throw some welcome rays on one of the darkest periods of history in the valley of the Tigris and Euphrates. But first of all, they again have brought vividly and impressively before our eye’ the one fact that Babylonian civilization did not spring into existence as a deus ex machina ; that behind Sargon I and Naram-Sin there lies a long and uninterrupted chain of development cov- ering thousands of years; and that these two powerful rulers of the fourth millennium before Christ, far from leading us back to “the dawn of civilization,” are at the best but two prominent figures from a middle chapter of the early history of Babylonia. 1A vase of ordinary gray pottery, 23 cm. high, was found 7.40 m. below this pavement ‘‘ directly beneath the line of the very ancient curb, and near to a perpendicular let fall from the EH. corner of the altar.’? The stratum which produced this vase, according to Haynes, ‘‘ was literally filled with potsherds of small size and generally brick red in color’’ (Report of Sept. 14, 1895). 242 OLD BABYLONIAN INSCRIPTIONS Il. THE INSCRIBED MONUMENTS OF SARGON’S PREDECESSORS. Although more than 500' mostly fragmentary antiquities of Sargon and his predecessors have been excavated in Nuffar, it may at first seem strange that nearly all of them were discovered out of place, above the platform of Ur-Gur. But if we examine the details more closely, we will easily find the explanation of this remarkable fact. Almost all these monuments that, on the basis of strong paleographic evi- dence and for various other reasons, must be ascribed to this early phase of Babylo- nian history,” were found in a stratum on the §.-E. side of the ziggurrat, between the facing of the latter and the great fortified wall which surrounded the temple. This stratum varies in thickness. ‘In some places it lies directly upon the crude brick pavement of Ur-Gur, while in other places it reaches a height of c. 1 m. above this platform.”*® Few of the objects found were whole, the mass of them was broken and evidently broken and scattered around on purpose. Most of the fragments are so small that during the last three years it needed my whole energy and patience, com- bined with much sacrifice of the eyesight, to restore the important inscriptions pub- lished on the following pages (particularly Pls. 36-42). The apparent relation in which this stratum stands to a peculiar building in its immediate neighborhood will furnish the key to the problem. AN ANCIENT TEMPLE ARCHIVE. Directly below the great fortification wall of the temple to the S.-H. of the zig- gurrat, Mr. Haynes discovered recently a room 11 m. long, 3.54 m. wide and 2.60 m. high. It showed nowhere a door or entrance in its unbroken walls, and there can be no doubt “that the room was a vault entered by means of a ladder, stairway or other perishable passage from above.” This structure “was erected on the level of Naram-Sin’s pavement,” and yet it was made of the same bricks which compose the ‘Stamped bricks being excluded. 2Cf. proof below. ® Haynes, Report of Dec. 14, 1895. CHIEFLY FROM NIPPUR. 243 body of Ur-Gur’s ziggurrat and platform. How is this discrepancy to be explained ? By the simple assertion, suggested already by the absence of a door in the walls of the building, that the room was underground, a cellar reaching from the top of Ur-Gur’s platform down to the level of Naram-Sin’s pavement.’ The access from above being on the Ur-Gur level, it is clear that the vault was built by this king himself. Our interest in the unearthed building is still increased by the discovery of another smaller? room of exactly the same construction and material below it. Separated from the later vault by a layer of earth and débris 60 cm. deep, it lies wholly below the level of Naram-Sin’s platform. In its present form this lower cellar cannot, however, antedate Sargon, nor was it built by this king himself or by his immediate successor. From the fact that the bricks of both rooms are identical “in size, form and general appearance,”* and that a brick stamp of Sargon was discovered beneath the founda- tions of the lower walls, we draw the following conclusions: (1) At the time of Sargon a cellar existed at this very spot, as indicated by the presence of his stamp below the level of his dynasty ;* (2) Ur-Gur found and used this cellar, but rebuilt it entirely with his own bricks. And as he raised the foundation of his ziggurrat far above the old level, he also raised the walls of the old chamber to the height of his new platform. (3) For some unknown reason—probably because the pressure of the neighboring temple fortifications from above, together with the yearly rains, the principal enemies of Babylonian sun-dried brick structures, had ruined the vault °—he changed its foun- dation afterwards and laid it on a higher level, at the same time widening the space between its two longer walls. It can be easily proved that this underground building was the ancient storeroom or archive of the temple. ‘ A ledge c. 0.5 m. wide and 0.75 m. above the floor extended entirely around the room, serving as a shelf for the storage of objects in due form and order.”’ ‘* A circular clay tablet together with two small tablets of the ordinary form and. five fragments were found on it,” and five brick stamps without handles were lying within its walls. And finally a similar room filled with about 30,000 clay tab- lets, inscribed pebbles, cylinders, statues, etc., was discovered by de Sarzec, 1894, in a 1The height of its walls agrees with the distance between the tops of Ur-Gur’s and Naraém-Sin’s platforms. *It is only 2.15 m. wide, and the walls are 92 cm. high in their present ruined condition. 3 Haynes, Report of Dec. 14, 1895. *Cf. above, p. 235, note 2. °On this theory it can be easily explained why a few tablets were found on the ledge of the lower room and brick stamps without handles were discovered on the floor of the same room. 6 Haynes, Report of Dec. 14, 1895. This ledge existed in both chambers. It was built up with the walls and consisted of crude bricks capped by a layer of burned bricks (Report of Dec. 21, 1895). TIn the lower vault (Haynes, Report of Dec. 21, 1895). In the midst of this lower chamber was ‘‘a hemispheri- cal basin of pottery set in a rim of stone,’’ the original use of which is still unknown (Report of Dec. 14, 1895). 244. OLD BABYLONIAN INSCRIPTIONS small mound at Tello,’ by which the true character of our building is determined be- yond question. The French explorer was more fortunate than Mr. Haynes in finding his archive undisturbed, but it will always remain a serious loss to science that the contents of the archive of Tello could not have been saved and kept together.’ The vault of Nippur had been robbed by barbarians of the third millennium before Christ, as I infer from the following facts and indications: 1. Nearly all the objects above referred to were excavated from a well-defined stratum in the neighborhood of this storeroom. From the position in which they were found, from the fact that none, except door-sockets in diorite, were whole, and from the extraordinarily small size of most fragments, it becomes evident that the contents of the archive were broken and scattered intentionally, as previously stated. 2. Three of the rulers of the dynasty of Isin built at the temple of Nippur,’ and an inscribed brick of Ur-Ninib was found among the fragments recovered from this stratum. ~- It is therefore clear that the destruction of the vases, brick stamps, etc., did not antedate Ur-Ninib’s government. As no document later than his time has been rescued from this stratum, it is also manifest that the deplorable disaster occurred not too long after the overthrow of his dynasty. 3. The archive existed however as late as the second dynasty of Ur. For Bur- Sin II wrote his name on an unhewn block of diorite, presented to Bél many centuries before by Lugal-kigub-nidudu, a pre-Sargonic' king of Ur and Erech, and turned it into a door-socket for his own shrine in Nippur.’ That the archive could not have been de- stroyed in the brief interval between Ur-Ninib and Bur-Sin II, so that the latter might have rescued his block from the ruins, results from a study of the general his- tory of that period, however scanty our sources, and of the history of the city of Nip- pur at the time of Ine-Sin, Bur-Sin II and Gimil (Kat)-Sin’ in particular. All the 10f. Heuzey, Revue d’ Assyriologie III, pp. 65-68. The description of this archive chamber excavated in Tello may find a place here: ‘‘Ces plaquettes de terre cuite, réguli¢érement superposées sur cing ou six rangs d’épaisseur, remplissaient des galeries étroites, se coupant 4 angle droit, construites en briques crus et garnies des deux cétés de banquettes, sur lesquelles s’étendaient d’autre couches de semblables monuments. Les galeries formaient deux groupes distincts, mais voisins l’un de l’autre.”’ 2 The thievish Arabs seem to have scattered their rich harvest everywhere. So far, I have examined about 2000 of these tablets myself. But not less than c. 10,000 have been offered to me for sale by dealers of Asia, Europe and: America within the last year. They all come from Tello. Cf. [lilprecht, Recent Research in Bible Lands, p. 80. ° Cf. Part I, pp. 27 f. and above, p. 230, note 1. *For the proof of this statement cf. below. 5Cf. Pl. 18, No. 21, and Part I, ‘‘ Table of Contents,”’ p. 49. Bur-Sin II repeated only what had been done by Sargon I long before. Cf. Part I, ‘‘Table of Contents,” p. 47 (No. 1), and below. 6 That Gimil-Sin was the direct successor of Bur-Sin II follows from P]. 58, No. 127, and that Ine-Sin was the im- mediate predecessor of Bur-Sin was inferred by Scheil from a contract tablet (Recwei] XVII, p. 38, note 3). The men- tion of the devastation of Shashru on this Tello tablet is only of secondary importance in itself, as the same event CHIEFLY FROM NIPPUR. 245 three kings mentioned devoted their attention to the interests of Inlil and Ninh and other gods worshiped in Nippur, as we learn from excavated bricks and door-sockets (Pl. 12 f.),’ from two chronological lists (Pl. 55, No. 125, and Pl. 58, No. 127),’ and from the large number of dated contracts discovered in Tello, Nuffar and other Babylo- nian mounds.’ ‘That the country as a whole was quiet and enjoyed peace and prosper- ity under their government, is evident from the many business contracts executed everywhere in Babylonia and from certain statements contained in them. The con- stant references to successful expeditions carried on by Ine-Sin against the countries of Karhar", Harshi*, Simurrum",' Lulubu", Anshan™ and Shashru“, by Bur-Sin II occurred at other times (e. g., in Bur-Sin’s sixth year, Pl. 58, No. 127, Obv. 6). But the fact that this conquest is placed between Bur-Sin’s accession to the throne and a very characteristic event at the close of Ine-Sin’s govern- ment (cf. Pl. 55, No. 125, Rev. 18-21) settles the question. Ine-Sin ruled at least forty-one years, according to the chronological list on Pl. 55. As, however, a part of it is wanting, it will be safe to assign a reign of c. 50 years to him. Bur-Sin II ruled at least twelve years (P1.58, No. 127), and in all probability not more than sixteen to eighteen Years. That the events mentioned on the two tablets are arranged chronologically, is beyond question. For (1) events which happened more than once are quoted in their consecutive order, but often separated from each other by other events which occurred between them, Cf. Pl. 55, Rev. 3 and 10; Rev. 4, 5 and 11, and especially Oby. 5 and Rev. 15 (between the two similar events lie twenty-eight years!). (2) In casea year was not characterized by an event prominent enough to give it its name, such a year is quoted as ‘‘joined to”’ or “‘ following” the previous year in which a certain event took place (ush-sa). Cf. Pl. 55, Rev. 7-8, 11-12, 13-14, 16-17, 18-20. (8) As we expect ina list arranged chronologically, Pl. 58, No. 127, opens with ‘“‘the year in which Bur-Sin became king.” If the king accomplished something worth mentioning in the year of his accession, this deed was added. Cf. Pl. 58, No. 127, Rev. 4: Mu dingir Gimil- dingir Sin lugal Urumki-ma-ge ma-da Za-ap-sha-liki mu-gul-a ‘In the year when (Gimil-Sin became king and =) King Gimil-Sin brought evil upon the land of Zapshali.’’ 1Cf. also Peters in The American Journal of Archeology X, p. 16 f. 2 Cf. No. 125, Obv. 2, 4, 10, 17, 18 (Ine-Sin), No. 127, Obv. 3, Rev. 3 (Bur-Sin II). 3 Cf. for the present Scheil in Recueil XVII, p. 37 f. *Ona tablet in Constantinople written at the time of Ine-Sin, we read the following date: mw Simu-ur-ru-umki Lu- lu bukiba-gul. From the fact that Simurru and Lulubware here mentioned together, Scheil (Recuedl XVII, p. 38) draws the conclusion that ‘‘Simuru se trouvait donc dans les mémes parages que 14 ot la stele de Zohab fixe le pays de Lulubi.” This assertion is by no means proven. The king may have conquered two countries far distant from each other in the same year. I call attention to Scheil’s theory in order to prevent conclusions similar to those which for several years were drawn from the titles of Nebuchadrezzar I (col. I, 9-11: sha danna matu Lulubi ushamkitu ina kakki, kashid matuAmurri, shalilu Kashshi) and led to curious conceptions about the land Amurri (ef. e.g. Eduard Meyer, Geschichte des Alterthums, p. 329, and especially Winckler, Untersuchungen, p. 37, note 2). Hommel’s identification of Simurru with Simyra in Phenicia is by far more probable (Aus der babylonischen Altertumskunde, p. 9). 5 Pl. 55, No. 125, Rev. 8; resp. Rev. 6, 10; resp. Rev. 4, 5, 11 ; resp. Scheil, 7. c., p. 837 (beginning); resp. Rev. 18; resp. Rev. 21. In connection with Anshan it may be mentioned that Scheil in Recueil XVII, p. 38 (especially note 6), translated Pl. 55, No. 125, Rev. 9: mu dumu-sal lugal pa-te-si An-sha-anki-ge ba-tug by ‘‘année ow la fille du roi devint patesi dans le pays d’Anshan.’’ Notwithstanding that Hommel (Aus der babylonischen Altertumskunde, p. 9) and Sayce (in The Academy of Sept. 7, 1895, col. b) reproduce this translation, which grammatically is possible, I reject it on the ground that there is no evidence that in ancient Babylonia women were permitted to occupy the high- est political or religious positions independently, and translate : ‘‘In the year when the patesi of Anshan married a daughter of the king (tug = aldzu, “to take a wife, to marry,’”’ cf. Delitzsch, Assyrisches Handworterbuch, p. 42). Ns By SOW, SQV00, Z}I9), 246 OLD BABYLONIAN INSCRIPTIONS against Urbillum™, Shashru™ and Rite-tar(?)hu",’ and by Gimil (KA4t)-Sin against Zapshali", testify to the same effect. Moreover, a number of other tablets which belong to members of the same dynasty, but cannot yet be referred to definite kings, mention Avmash", Humurtt”’ and Huhu(nu)ru™* as devastated or invaded by Babylo- nian armies.’ Several of these cities and districts were situated on the east side of the Tigris and must be sought in Elam and its neighboring countries. We begin now to understand why the Hlamites soon afterwards when they invaded Babylonia made such a terrible havoc of the temples and cities of their enemies; they simply retaliated and took revenge for their own former losses and defeats. 4. When the Cassite kings conquered Babylonia, the site of the ancient archive chamber was long forgotten and buried under a thick layer of débris. Their own store- room, in which all the votive objects published on Pls. 18-27 and Pls. 60 f., Nos. 133- 142, were discovered, was situated at the edge of a branch of the Shatt-en-Nil outside of the great S.-H. wall of the temple of Bél.? The destruction of the archive under discussion must therefore have taken place between the oveithrow of the second 1 Pl. 58, No. 127, Obv. 2; resp. Obv. 6; resp. Obv. 7. ?Pl. 58, No. 127, Rev. 4. * Cf. Scheil, 1. ¢., p. 88. The city of Marhashi (in N. Syria, according 1o Hommel, J. c., p 9) is mentioned in con- nection with a daughter of Ine-Sin on Pl. 55, No. 125, Obv. 14. “In view of all these facts above mentioned, Homme! will doubtless change his view (that the kings of the second dynasty of Ur ‘‘were apparently confined to this city, as they did not possess Sumer and also lost Akkad’’). That they were not confined to Ur, but possessed the whole south is proven by their buildings in Eridu (I. R. 3, No. XII, 1, 2) and in Mippur (cf. also the statements of the two chronological lists). If Winckler’s theory as to the seat of the shar7 ut kibrat irbittt was generally accepted (Hommel apparently does not accept it), the second dynasty of Ur by this very title would also have claimed N. Babylonia. Whatsoever our position may be as to the meaning of this and other titles, as a matter of fact, the kings of the second dynasty of Ur possessed the south of Babylonia, and it is impossible to believe that kings who were the lords of 8. Babylonia and conquered parts of Arabia, Syria, Elam and other dis- tricts between the four natural boundaries defined in Part I, p. 25, note 4, and who doubtless in consequence of their conquests assumed the proud title “king of the four quarters of the world,”’ should not have been in the possession of all Babylonia (the case of Gudea is entirely different). Thekings of the second dynasty of Urchanged the title of their predecessors, not because they had lost Sumer and Akkad, but because they owned more than the old title indicated. The title of Sumer and Akkad—as I understand its meaning—is practically contained in that of ‘‘king of the four quarters of the world”’ (Part I, pp. 24 f.), and the kings of the second dynasty of Ur dropped it therefore for the same reason as Dungi, when he assumed the title shar kibrat arba’im (Z. A., 111, p. 94). As to the meanings of the different titles, Hommel (whose latest opinion is briefly stated in 4us der babylonischen Altertumskunde, p. 8) and I agree entirely, differing from Winckler esrecially in his interpretation of shar kibrat arba’im and shar mdatuShumeri u Akkadi in the oldest Babylonian inscriptions down to Hammurabi. Notwithstanding that, or rather because I read and stucied his Altorientalische Forschungen ILL, pp. 201-248, and all his previous papers on the same subject sime ira et studio again and again, I have been unable to convince myself of the correctness of his views. Tiele (Z. A., VII, p. 868), Lehmann (Shamashshumukin, pp. 68 ff.), Hommel (J. c.) and I apparently reached similar conclusions on this important question. 5 Cf. Part I, ‘‘Table of Contents,” p. 48 (Pl. 8, No. 15). Cf. also Peters in The American Journal of Archeology X, p. 15. CHIEFLY FROM NIPPUR. 2AT. dynasty of Ur and the beginning of the Cassite rule in Babylonia. The history of the temple of Bél during this period is enveloped in absolute darkness. No single monu- ment of the members of the so-called first and second Babylonian dynasties has yet been excavated in Nuffar. Apparently our temple did not occupy a very prominent place during their government. And how could it be otherwise? Their rule marks the period of transition from the ancient central cult of Bél in Nippur to the new rising cult of Marduk in Babylon. Bél had to die that Marduk might live and take his place in the religious life of the united country. Even the brief renaissance of the venerable cult of “the father of the gods” under the Cassite sway did not last very long. It ceased again as soon as the national uprising under the dynasty of Pashe led to the overthrow of the foreign invaders, who had extolled the cult of Bél at the expense of Marduk in Babylon,’ and to the restoration of Semitic power and influence in Babylonia, until under the Assyrian kings Esarhaddon and Ashurbinapal a last attempt was made to revive the much neglected temple service in the sanctuary of Nippur. : 5. The breaking and scattering of the vases point to a foreign invasion and to a period of great political disturbance in the country. No Babylonian despot, however ill-disposed toward an ancient cult, and however unscrupulous in the means taken to suppress it, would have dared to commit such an outrage against the sacred property of the temple of Bel. In all probability therefore the ancient archive chamber of the temple was ransacked and destroyed at the time of the Elamitic invasion (c. 2285 B.C.), when Kudur-Nankhundiand his hordes laid hands on the temples of Shumer and Akkad. That which in the eyes of these national enemies of Babylonia appeared most valu- able among its contents was carried to Susa’ and other places; what did not find favor with them was smashed and scattered on the temple court adjoining the storehouse. From the remotest time until then apparently most gifts had been scrupulously pre- served and handed down from generation to generation. Only those movable objects which broke accidentally in the regular service, or which purposely were buried in con- nection with religious rites, may be looked for in the lowest strata of Ekur. AGE OF THE INSCRIBED MONUMENTS. Having explained why the most ancient documents so far excavated in Nuffar were found in pieces above the platform of Ur-Gur’s ziggurrat, I now proceed to determine the general age of these antiquities and their relation to the inscriptions of Sargon I. 1@f. Part I, pp. 30 f. 2Cf. Part I, p. 31. 248 OLD BABYLONIAN INSCRIPTIONS The inscriptions Nos. 86-112 have many palzographie features in common and doubt- Jess belong to the same general period, the precise extent of which cannot be given. Two groups, however, may be clearly distinguished within it, differing from each other principally in the forms used for mu (Briinnow, List 1222) and dam (ibid., 11105). Instead of the two familiar Old Babylonian characters, in mu the two pairs of parallel lines found at or near the middle of the horizontal line, sometimes cross each other (Nos. 92, 5; 98, 3; 99,4; 101, 3, ete.), while dam occasionally has a curved or straight line between the two elements of which it is composed (No. 111, 3 and 6; No. 98, 2 and 5; cf. No. 94, 3).' This peculiar form of dam has so far not been met with outside of a very limited number of inscriptions from Nippur; that of mw occurs also on the barrel cylinder of Urukagina, although in a more developed stage. Whenever one of these characters has its peculiar form in an inscription of Nippur, the other, if accidentally occurring in the same inscription, also has its peculiar form as described above (cf. No. 94, 3 and 4; No. 98, 2(5) and 3; No. 111, 3 and 6). The two char- acters represent therefore the same period in the history of cuneiform writing, to the end of which the cylinder of Urukagina also belongs. This period has not yet been definitely fixed. As various historical considerations seemed unfavorable to placing this ruler after the other kings of Shirpurla, Jensen provisionally placed him before them ;* Heuzey was less positive ;* Hommel® and Winckler® regarded him as later, while Mas- pero, without hesitation, but without giving any reasons, made him “the first in date of the kings of Lagash.”’ Aside from the reasons given by Jensen, and a few simi- lar arguments which could be brought forth in favor of his theory, the following palx- ographic evidence proves the chronological arrangement of Jensen and Maspero to ke correct : 1. The peculiar form of mw occurs in inscriptions from Nippur which, if deter- 1 This short line, about the significance of which I refer to my greater work, Geschichte und System der Keilschrift, was originally curved, became then straight and was later placed at the end of the character (No. 93, 6; 96, 4; 113, 12), finally developing into a full-sized wedge (De Sarzec, Découvertes en Chaldée, Pl. 26, No.1, col. II, 1; Heuzey in Revue d’ Assyriologie II, p. 79, No. 1, 13 [a duplicate of this inscription is in M. I. O., Constantinople], and the present work, No. 123, Obverse, 1). Sometimes this line is entirely omitted (No. 112, 6). ? De Sarzec, J. c., Pl. 32, col. I, 7; col. II, 1, 4, 12; col. Ill, 3, 7. The form of mw is more developed in Uruka- gina’s inscription, indicating that the latter is somewhat later than the corresponding Nippur texts. On the other monuments of Urukagina the regular Old Babylonian form is used exclusively. %In Schrader’s Ketlinschriftliche Bibliothek, Vol. III, Part 1, p. 8. * Formerly he regarded him as decidedly later than the other kings of Lagash (in De Sarzec, Découvertes en Chal- dée, pp. 110, 112). More recently he expressed himself as doubtful : ‘‘Il en résulte que le roi Ourou-ka-ghi-na doit étre tenu, soit pour appartenir & une dynastie antérieure a celle du roi Our-Nina, soit pour avoir, apiés Vapparition des premiers patési, relevé le titre royal a Sirpourla’’ (Reoue d’ Assyrdologie II, p. 84). 5° Geschichte Babyloniens und Assyriens, pp. 290f. 5 Geschichte Babyloniens und Assyriens, p. 41. 1 The Daun of Civilization, p. 604. CHIEFLY FROM NIPPUR. 249 mined by the character of dam alone, must be classified as older than the royal in- scriptions of Tello. 2. The form of mw employed in Urukagina’s cylinder does not occur in any other inscription of Tello. The cylinders are therefore to be regarded as older than the other monuments, if it can be shown that this peculiar form of mu represents a more ancient stage of writing’ and did not originate from an accidental prolongation of certain lines in mu by a careless scribe.” 3. The very pronounced forms cut in stone vases (as, e. g., found in No. 98, 3; 101, 4; 92, 5, and first of all in No. 94, 4) force us to eliminate the element of acci- dent. But, besides, it can be proved by an analysis of the character mw itself that the regular Old Babylonian sign is only a later historical development of a more ancient form. The correct interpretation of the original picture will, at the same time, enable us to catch an interesting glimpse of certain prehistoric conditions in ancient Shumer. According to Houghton,’ a close relation exists between the character for mu and hu (Brinnow, J. c., 2044) and the first part of the character for am (zbid., 2087). I trust no Assyriologist of recent date has ever taken this attempt at solving a palezeographic problem very seriously. ‘The sign for nam has no connection with the other two char- acters and is no compound ideogram, but, in its original form, represents a flying bird with a long neck.* Since in Babylonia, as in other countries of the ancient world, the future was foretold by observing the flight of birds, this picture became the regular ideogram for “fate, destiny ” (shimtu) in Assyrian. The original picture for mz, on the other hand, is no bird, but an arrow whose head formerly pointed downward, and whose cane shaft bears the same primitive marks or symbols of crossed lines as are characteristic of the most ancient form of arrow used in the religious ceremonies of the North American Indians.’ As the shaft was represented by a single line in Baby- 1This argument is conclusive, as the theory, according to which later writers occasionally imitate older forms of cuneiform (or linear) characters, in the sense generally understood by Assyriologists, is without any foundation and against all the known facts of Babylonian paleeography. Cf. my remarks in Part I, pp. 12f. 2 Jensen’s hesitation, so far as founded upon the form of the character ka, can be abandoned, as the form of this character is surely far older than Gudea. 3In the Transactions of the Society of Biblical Archeology VI, pp. 464f. 4 This fact becomes evident from a study of the oldest forms in the inscriptions of Telloand Nippur. The original picture is still found on the most ancient Babylonian document in existence, unfortunately scarcely known among Assyriologists. It is (or was) in the possession of Dr. A. Blau and was published by Dr. W. Hayes Ward in the Proceedings of the American Oriental Society, October, 1885. The bird represented is therefore no ‘‘swallow’’ (Hom- mel, Sumerische Lesestiicke, p. 6, No. 67), but a large bird with a long neck, such as a goose or a similar water bird found on the Babylonian swamps. Later our picture was also used as the ideogram for ‘‘swallow,”’ designating her as the flying bird par excellence, as the bird nearly always in motion when seen at day time. 5 As I learned through the courtesy of Mr. Frank Hamilton Cushing of the Bureau of American Ethnology in the Smithsonian Institution at Washington. After a correspondence on this subject it became evident that we had 250 OLD BABYLONIAN INSCRIPTIONS lonian writing, the original mark carved upon its surface had to be drawn across it. Instead of 2 oa we find, therefore, XO from which, by short- ening the crossed lines, the regular form > developed at a later time. The correctness of this explanation is assured by the otherwise inexplicable absence of an ideogram for ussw, “arrow,” in Assyrian. For it is impossible to conceive that a people using the bow in their system of writing should have altogether excluded the arrow, which played such a conspicuous réle in the daily life and religious ceremonies of ancient nations in general. But how is it to be explained that our ideogram does not mean “arrow ” at all, but signifies “name?” Just as the picture of a flying bird in writing proper was used exclusively with reference to its religious significance, in order to express the abstract idea of “fate, destiny,” so the arrow with the marks or symbols of ownership (originally two crossing lines’) carved on the shaft became the regular ideogram for “personality” or “name.” ‘The same association of ideas led to exactly the same symbolism and usage among the North American Indians, with whom “the arrow” is the symbol of personality.” It becomes now very evident that the Babylo- nian seal-cylinder, with its peculiar shape and use, has developed out of the hollow’ shaft of an arrow marked with symbols and figures, and is but a continuation and elaboration in a more artistic form of an ancient primitive idea. From palzeographic and other considerations it is therefore certain that Urukagina lived before the ancient kings of Shirpurla, while the inscriptions published in the present work as Nos. 90, 91, 92, 94, 98, 99, 101, 111 are still older than Urukagina. The interval between him and the following rulers of Tello who style themselves “kings ” cannot have been very great, however. They all show so many paleographic features in common that they must be classified as an inseparable group. ‘To the both reached the same conclusions as to the oldest form and significance of the arrow in picture wriling by pursuing entirely different lines of research. My arguments, corroborated by Mr. Cushing’s own investigations and long resi- dence among tribes which still practice many of the ancient primitive rites and customs, become therefore conclusive in regard to the original form of the character mu. I quote from Mr. Cushing’s letter the interesting fact that the above drawn arrow with two pairs of crossing lines on its shaft is called by the Zuni a'thlua ‘speeder (commander) of all’’ (namely, of all the other arrows used in their religious ceremonies). A treatise on the ceremonial use of the arrow among the Indians, by Mr. Cushing, is in press. 1 Still used with the same significance in Europe and America by persons who cannot write, if they have to affix their names to legal documents. The crossed lines on the Indian arrows have a deep religious significance, according to Cushing. > Cf. on this whole subject Culin, Korean Games, pp. XXIf. To Prof. Dr. Brinton and Mr. Stuart Culin I am indebted for recent information on this subject. 5 Because made of bulrushes, growing abundantly along the marshes and canals of lower Babylonia. CHIEFLY FROM NIPPUR. 251 same age doubtless belong most, if not all, of the other inscriptions published on Pls. 36-47 (No. 112). I shall prove my theory in detail by the following arguments: I. Palzographically they exhibit most important points of contact with Uruka- gina, Ur-Nina, Edingiranagin, Enanatuma I, Entemena, Hnanatuma II, especially with the first three mentioned. a. Characteristic signs are identical in these Nippur and Tello inscriptions. Cf, e.g. gish, No. 87, col. I, 10, col. II, 37, No. 110, 4 f. e., with the same sign in the texts of Ur-Nina and Edingiranagin ;' ban, No. 87, col. I, 10, col. II, 37 (ef. No. 102, 2) with the same sign in the texts of Edingiranagin; a, No. 86, 8 (Var.), 1 f.e., No. 87, passim; No. 96,2; No. 104,3; 106, 4; 110, 8f.e, 112, 7, with the sign used by Ur-Nina, Edingiranagin, Hnanatuma JI, Entemena (ef. also the present work, No. 115, col. I, 7, col. 11, 1, 2, ete.); shu, No. 87, col. III, 34 (and Var.) with Uru- kagina, Edingiranagin; da, No. 86, 7, No. 87, col. I, 19, col. I, 18, 20, 29, ete., with the sign used by Ur-Nina, Edingiranagin, Entemena; a (ID), No. 87, col. II, 41 (Var.) with Entemena (No. 115, col. I, 5); ta, No. 87, col. I, 46, col. IT, 4, 12, with the same sign used by Urukagina, Ur-Nin4, Edingiranagin, Entemena; ma, No. 88, col. IIT, 2, with the same sign used by Urukagina, Endigiranagin;’ ma, No. 87, col. II, 40 ff, with the same sign used by Urukagina, Hdingiranagin; and many other characters. b. The script is almost entirely linear like that of Urukagina,*® Ur-Nin& and Edingiranagin. c. They show certain peculiarities in the script, which so far have been observed only in the most ancient texts of Tello: (1) Lines of linear signs running parallel to a separating line (marking columns and other divisions) frequently fall together with this latter so that the character now appears attached to the separating line above, below, to the right or left. SSometimes characters are thus attached to two sep- arating lines at the same time. Cf. No. 87, col. I, 5 (ma), 12 (ka), col. II, 9 (shu), 17 (la), 29 (2), col. III, 36 (wr), No. 106, 2 (27n), and many others written on different fragments of No. 87.' (2) In accordance with this principle two or more characters 1Tn these quotations, as a rule, I shall abstain from giving the exact passages, as I expect that everybody who examines my arguments has made himself familiar with the paleography and contents of the most ancient inscriptions of Tello before, and to those who have not done so, I do not intend to give introductory lessons in the limited number of pages here at my disposal, in fact for those I do not write. 2 Also used by Naram-Sin, cf. No. 120, col. I, 4. ‘Except of course his barre] cylinder, which has cuneiform characters, as it was inscribed with a stylus. ‘For this palzographic peculiarity in the inscriptions of Tello, cf. Urukagina (De Sarzec, Décowvertes, Pl. 82, col. II, 9, 10, col. III, 2, 5, col. IV, 3, 9, col. V, 2, 4); Ur-Nima (De Sarzee, J. c., Pl. 2, No. 2, col. I, 1, 3, Revue d’ As- syriologie II, p. 84, 3 and 4; p. 147, col. I, 3, 5, col. III, 3, 6, col. IV, 3, 5); Edingiranmagin (De Sarzec, I. ¢, Pl. 4, Frag. A, col. I, 6, col. I, 3, 4, 5, 10, etc.; Pl. 31, No. 2, col. I, 1-4, 6, col. II, 1-8, 5, ete.); Enanatuma I (Reoue 252 OLD BABYLONIAN INSCRIPTIONS standing in close proximity to each other frequently enter into a combination, forming so-called ligatures.| Cf. No. 86, 5 Var. (md-na), 8 (tab-ba, cf. also Variants), 15 Var. (ki-gub); Part I, Pl. 14, 2 (du-du) ; No. 87, col. II, 9 (ma-shu), 20 Var. (da- ga), 34 (ki-ag), 45 (da-gz, cf. Var. gi-gz),” col. III, 21 (ba-dag),’ 34 (PA [first half of the character stb |'-gal); No. 93, 7 (Shul-pa); No. 94,1 (Min-din-dug (?) );> No. 98, 2 (dam-dumu); No. 111, 6 (na-da).’ On the monuments of Tello this tendency to unite two characters into one is almost entirely confined to the inscriptions of Ur- Nina.* The best illustration is afforded by the writing of the name of his son, Nina- shu-banda. The four signs which compose the name are contracted into one large sign, the earliest example of a regular monogram in the history of writing (De Sar- zee, l. c., bl. 2°", No. 1). A number of signs which occurred always’ in the same @ Assyriologie II, p. 31, 1-5, 9, 11, 14 f.); Entemena (De Sarzee, J. ¢., Pl. 5, Nos. 2, 4°and 5; Pl. 31, No. 3, col. I, 2,4, 5, col. Il, 3 f£ ; Revue d’ Assyriologie II, p. 148, col. I, 1-6, etc.) ; Knanatuma II (De Sarzec, J. ¢., Pl. 6, No. 4, 2-5, 7f.) For other examples of Entemena’s text in the present work, cf. Nos. 115-117. Apparently Dr. Jastrow had not seen a Tello inscription when he wrote his remark in Z. A. VIII, p. 217. 1Tn a limited measure the same peculiarity occurs in several Assyrian inscriptions, c. 3000 years later. Cf., e. g., ¢-na, in the inscription of Tiglathpileser I (I #., 9 ff), ira pa, Salm. Obdel., 1. 160, 176 (Hilprecht, Assyriaca, p. 27, note), etc. 2Col. II, 48. ki-nin Unugki-ga, 44. ganam-gad-shakir-a-dim, 45. shig mu-da-gi-gi. The last character in 1. 38, which remained unidentified for such a long time (cf. Amiaud et Méchineau, Tableau Comparé, No. 122, Jensen in Schrader’s K. B. III, part 1, p. 16, note 4; Scheil in Recueil XV, p. 63; Hommel, Sumerische Lesestiucke, p. 32, No. 376) is identical with Briinnow, List 5410. It has in the ancient inscriptions the two values ga and ma (for the latter cf, e. g., No. 87, col. IL, 19 (kalam-ma), 29 (Urum*i -ma) ). On PI. 50, col. II, 4, read NA-GA = ishkun (and col. III, 4f., KI-GAL (= kigalla) ish-pu-uk, against Scheil in Recuedl XV, 62 f.). 8Col. Ill, 19. nam-ti-mu, 20. nam-ti, 21. ga-ba-dag-gi—‘‘ unto my life he may add life.”’ *PA-gal LU sag gud, read sib (PA-LU sag-guda-gal, ‘the shepherd having the head of an ox’’ = ‘‘the ox- headed shepherd,” a synonym of king, according to Jensen. 5On the god Shul pa-ud-du, cf. Jensen, Kosmologie, pp. 126 f., and in Schrader’s K. B., Ill, part I, p. 65, note 11 (Umun-pauddu). Oppert read Dun-pa-e. 6«The goddess who destroys life,’ an ideogram of Bau or Gula (Briinnow, Zis¢ 11084, ef. IIL R., 41, col. II, 29-31; III R., 48, col. IV, 15-18, and the present work, Pl. 67, col. III, 1-5). The same deity is mentioned No. 95, 1, No. 106, 1, No. 111, 1. On the value of dug cf. Hommel, Sumerische Lesestitcke, p. 5, No. 53, and p. 12, No. 145. TCf. No. 99, 5. 8Cf, Revue d Assyriologie II, p. 147, col. III, 6 and 7, col. V, 1, 3, 6. 9Cf. No. 87, col. I, 5, 40, 42, etc. The linear sign is composed of ¢ (canal) + gi (reed) and originally denotes a piece of land intersected by canals and covered with reeds (cf. No. 87, col. III, 29). The land par excellence with these two characteristic features was to the Babylonians their own country, which therefore was called by the oldest inhabitants Ki + e+ gi = Kengi, ‘the land of canals and reeds,’’ From this correct etymology of Kengi and its use in the earliest texts (bar bar Kengi, No. 87, col. II, 21, and Enshagsagana en Kengi, No. 90, 3) it follows that the name does not signify ‘‘low-lands”’ or “‘ Tiefebene”’ in general in the ancient inscriptions, which alone have to decide its meaning (against Winckler in Méttetlungen des Akademisch-Orientalistischen Vereins zu Berlin, 1887, p. 12), but that it is the geographical designation of a well-defined district, Babylonia proper. As, however, Babylonia and low- lands are equivalent ideas, Kengi could also be used in a wider sense for ‘“‘low-lands” (matu) in general, CHIEFLY FROM NIPPUR. 253 combination and served to express but one idea or object, were regularly contracted at this early time and became compound ideograms, e. g., kalama “ country,” gishdin' “wine,” ete. (3) Lines of linear signs which run parallel to a separating line are often omitted, even if the sign is not directly connected with this latter. Cf. No. 'The peculiar way in which it is written in the oldest inscriptions of Tello, leaves no doubt as to its composition (gish + din). The analysis of this ideogram by Pinches (Sign List, No. 76a = kash + din), accepted by Delitzsch (Assyrisches Handworterbuch, p. 354), Jensen (in Schrader’s K. B. III, part 1, p. 27, note 6), Hommel (Sumerische Lesestticke, No 180) and others, must therefore be abandoned. For examples cf. Edingiranagin’s inscription un- earthed in London (Proc. Soc. Bibl. Arch., Nov. 1890), col. IV, 3, 7, col. V, 3: gishdin zu-zu-a; or Gudea D (De Sarzec, l.c., Pl. 9): 6. Ma-ganki, 7. Me-lug-gaki, 8. Gu-biti, 9. kur Ni tugki, 10. gu gish mu na gal la-a-an, 11. ma gishru-a gishdin (ste!), 12. Shir-pur-laki-shu, 13. mu-nt-tum—‘“‘ Magan, Meluha, Gubi, Dilmun, each (a2) of which possesses every kind of tree, brought a ship (laden) with timber and wine to Shirpurla.” Jensen’s question (in Schrader’s K. B. Til, part 1, p. 13, note 12), as to what Amiaud may have read in Ur-Nin4’s inscription (De Sarzee, J. c., Pl. 2, No. 1, col. IV, 1-3, which Jensen left untranslated) is answered by referring him to the Gudea passage just translated, and to Revue d’ Assyriologie II, p. 147, col. V, 3-6, together with De Sarzee, J c., Pl. 2bis, No. 1 (lower section, charac- ters standing immediately before the king). Amiaud, however (in Records of the Pust2 I, p. 65), as well as Oppert (io Revue W@ Assyriologie Il, p. 147) and Heuzey (in Revue d’ Assyriologie IIL, p. 16, and Découvertes en Chaldée, p. 170) wrongly read gish din (notwithstanding the passage from Gudea just quoted, lines 6 and 10, where the two respective characters are very different from each other !) as gan (kan) finding the name of Magan in the first line. The passage reads rather: 1. ma gishdin, 2. kura-ta, 3. gu gish gal, 4. mu-tum (?)—‘‘a ship (laden) with wine he brought from the country which possesses every kind of tree.’’ We are now enabled to understand the full significance of Ur-Nina’s perforated bas relief (De Sarzee, J. c., Pl. 2bis) which remained obscure to Heuz2y in his treatise mentioned below. These bas-reliefs and incised slabs (cf. the present work, Pl. XVI, Nos. 37f.) did not serve ‘‘a maintenir dressés, sur des autels ou sur des massifs de briques, divers engins consacrés; aux dieux et particaliérement des masses d’armes votives’’ (Heuzey, Les Armoiries Chaldéennes de Sirpourla, pp. 11 f., cf. pp. 6f.). For they would have been too small and weak for such a purpose. The true facts are rather these : (1) They accompanied donations of any kind made to the temple. But while such donations were consumed in the interest of the temple service (cf. Hilprecht, Z. A. VIII, p. 191 f.) or decayed in time (buildings) or died (slaves), etc., these tablets were preserved in the temple as lasting memorials to their munificent donors and served at the same time to induce other worshipers to similar acts of piety. (2) The hole in the middle of the tablets served to fasten it, by the aid of a nail, in the wall or floor of the temple, possibly on the altar itself. (3) The scenes, objects and inscriptions on these tablets generally illustrate and describe the person and work of the donor in relation to his deity. Ur-Nina’s more elaborate votive tablets (of which the smaller is only an excerpt, cf. De Sarzec, J. ¢., Pl. 2bis, pp. 168-173), accordingly represent two sides of the king’s work undertaken in the service of his god. In the upper section he has the dupshig (= dupshikhu), the symbol of masons, upon his head (exactly as Nabopolasser describes himself in the present work, Pl. 33, col. II, 57 ff.), and is surrounded by his children and page (Da-ni ta ‘‘at his side’’ = ‘‘ page,’’ not ‘‘in his hand,’’—Oppertin Revue d’ Assyr- tologie ILI, p. 16, note 1). This picture illustrates the accompanying statement: ‘‘Ur-Nina, king of Shirpurla, son of Nigalnigin, built the temple of Ningirsu, built the abew-banda (cf. Jensen in XK. B. III, part 1, p. 18, note t+), buiit the temple of Nin&.’’ In the lower section the same king, seated and surrounded by his children and his chief butler (Sag-antug ‘’ he is the chief’’), offers a libation of wine. This picture illustrates the words standing below the cup, ‘“‘a ship of wine he brought from the country which possesses every kind of tree.’’? The inscription of the bas-relief published by Heuzey in Les Armoiries Chaldéennes de Sirpourla reads : 1. Lay (DU-DU =abélu “to bring,” nazdzu ‘to set up ap) 2. sanga (Briinnow, List 5989) may, 8. dingir Nin gir su-ka, 4, dingir Nin-gir-su, 5. E-ninnti-ra, 6. lag, 7. sanga (cf. the present work, No. 87, col. I, 30, and No. 113 ,5) dingir Nin gir-su-ka ge, 8..... kita, 9. mu-na-ta-ud-du, 10. GAG + GISH (not gisal, Hommel, Sum. Lesest., No. 205) ura-shu, 11. mu-na-gim—“ Gift of the high-priest of Ningirsu to Ningirsu of the temple Eninnti. The gift of the priest of Ningirsu he brought from ....and worked it intoa....” “Ae DB SeAVO, SOMMIG AGL Diy: OLD BABYLONIAN INSCRIPTIONS 86, 3 Var. (ra), 4 Var. (4), 5 Var. (na); No. 87, col. I, 4 (Unug), 14 and 20 Var. (dingir), 19 Var., col. 11, 87 Var., 45, III, 34 Var. (da), 40 Var. (kalama) ; col. 31, 31 Var. (gum) ; col. III, 2 (wm), 23, 41 Var. (a), 29 (ma), 37 Var. (nam), etc. Out- side of the Nippur texts this peculiarity is almost confined’ to the inscriptions of Ur- Nina. Cf, ¢.9., De Sarzec, 1. ¢., Pl. 2°", No. 2, upper section (da in the name of Ab- da), zbid. (Ur in the name of Ur-Nina), Levue d’ Assyriologie II, p. 147, col. V, 4. II. The paleeographic evidence brought forth is conclusive. Nos. 86, 87 and the other texts referred to above, show all the characteristic features of the inscriptions of Urukagina, Ur-Nina and Edingiranagin. But besides they exhibit a number of paleo- graphic peculiarities which are altogether absent from the inscriptions of Tello, and must be regarded as characteristic features of an earlier stage of writing. They will be treated in full at another place.* I confine myself here to a brief statement of the following fact. A number of signs have a form representing almost the original pic- ture, others have at least a more original form than the inscriptions from Tello, even those of Urukagina not excepted. Cf. sum (No. 87, col. I, 17, the ear of a corn, cf. also 1. 45), gi (abid., col. I, 3, a reed, bulrush)*, @ (cbid., col. I, 31 in egi-a, a tattooed forearm with hand),* bar (2bid., col. II, 21; No. 98, 4 (the skin of an animal or) a coarse rug),’ lah (dbid., col. 1, 21, water poured out, therefore, “to wash”),” ra (2bid., 7One example is found in a text of Entemena (ne, ef. Revue d’ Assyriologie 1, p. 149, col. IV, 2). The way in which Ur is written in the name of Urukagina (De Sarzee, . c., Pl. 32, col. I, 1), furnishes the key to the origin of this peculiarity. For details on this subject I refer to my Geschichte und System der Keilschrift, which has been in prepa- ration for the last nine years. *In advance I warn Assyriologists not to regard a fourth palszeographic peculiarity (so far confined to these Nippur texts) as a mistake of the scribes: (4) If two linear signs which are to be connected grammatically stand close together in writing, yet without touching each other, frequently one line of the second running parallel to a line in the first is omitted entirely and has to be supplemented from the first sign. Cf. No. 87, col. III, 37: la-mi (sic /), 39: aga-mi (sic !), 40 Var.: mu-na (sic!) ; No. 108, 3: ma@-na (sic /). *In order to obtain a clear conception of the original picture, this sign must not be turned to the left (as Hough- ton, 1. ¢., p. 473, ard others did). For it is a law in cuneiform writing “that the characters are all and always reversed in the same way ; what (originally) was the right hand side became (later) the top’’ (Bertin, J. ¢., p. 6). The triangle on the left of our picture does not represent the lower end of the stem of a reed, but rather its top or cob. Cf. the corresponding pictures on the Assyrian monuments published in Layard, 7he Monuments of Nineveh, Second Series, e. g., Pl. 12, No. 1 (reproduced by Maspero in 7he Dawn of Civilization, p. 561). *The crossed lines do not represent ‘‘an ornamented sleeve’’ (Bertin, J. ¢., p. 9), but marks of tattooing (cf. Berger, “‘ Rapport sur les tatouages Tunisiens,”’ in Revwe @’ Assyriologie III, pp. 83-41). The cuneiform sign without these marks means ‘‘side’’ (da) ; with them, it denotes him who is at somebody’s side for assistance ; he who has the same marks of tattooing upon his arm, therefore has become his ‘‘ brother.’’? The sign for shesh, ‘‘brother,”’ denotes a person as the second child of the same family, while the former expresses tribal relations represented by a common symbol. ° According to Oppert (Fapédition en Mésopotamie, Tome Il, p. 64) and Bertin (J. c., p. 8) an altar. Impossible ! It represents the skin of an aninal or better a coarse rug spread upon the ground for persons of rank (and images of deities) to sit upon ; in other words, it denotes the place of honor, in exact harmony with the custom prevailing in the tents of Arabia and Mesopotamia to-day. Lehmann (Shamashshumuhin, p. 122) is therefore correct in giving CHIEFLY FROM NIPPUR. 255 cols ame Var, col) 1M, Wa) War, Scanaleye “to saulle29(s2— horn). eS to) irri- gate”’),' lugal (cbid., col. I, 1-3, the sign shows the remnant of the original arm” Cf. also the ideogram zag (tbid., col. I, 3, 38, ete.), gur (chid., col. III, 42 Var.),’ Kish (No. 92,3; No 102, 3; 103, 4),‘ ag (No. 85, 11 and 14),’ and many others for whose explanation I must refer to my Gleschichte und System der Keilschrift.’ All the stone inscriptions of Urukagina have the regular Old Babylonian sign for mu,’ just as the Nippur texts here treated. On the other hand, the Nippur texts have a large number of far more original forms of signs than the Urukagina and Ur-Nin% inscriptions published.* In view of these facts I can only draw one conclusion—that most of these Nippur texts are older than those of Urukagina. II. Another important fact corroborates my determination of the age of these ” 2 to bava(g) the original meaning, ‘‘seat,’’ instead of ‘‘chamber.’”’ This sign occurs frequently in the contracts of Nuffar (in a much more developed form) and was identified with bar by Scheil independently of me. Cf. Recwetl XVIL, p. 40d. 6 Suk(k)allu denotes the servant (gal) who pours out (sz) [namely water over his master’s hands and feet]. A word with similar meaning (zw) is apparently contained in zu-ad, ‘‘ocean,” which Hommel translated half correctly “house of water (?),”’ cf. Sumerische Lesestitcke, No. 6. Originally zw and sw had the same ideogram, which repre- sents a vessel (cistern?) into which water flows. Zu means, therefore, ‘‘to flow into,’’ or trans., “‘to pour into, to add,” then figur., ‘‘to increase one’s knowledge, to leara, to know.’? Zu-ab denotes ‘‘the house (abode) into which all the waters flow.’’ Swkkallu may be translated ‘‘chamberlain’’ (Kammerer), later it received a more general meaning. 1 Oppert already recognized the general significance of the picture (J. c., p. 64). But the exact analysis of the compound ideogram, which I discovered long before we excavated in Nuffar, remained obscure to him, Houghton, Sayce (Zransactions of Svc. Bibl. Arch. VI, p. 475) and others. Cf. a very curious form, which is but a mutilated “‘rq,’’ in col. I, 37, second Var. 2 The two elements lw ++ gal appear separated in No. 86, 2 Var., 13 Var.; No. 104, 7; No. 195, 7. ’ Successfully analyzed by Ball in Proc. Soc. Bibl. Arch. XV, p. 49. The line which continues beyond the head is, however, no continuation of the forearm, but represents the cushion between the head and the vessel upon which the latter rests. Originally the arm reached further to the rim of the vessel, as in the corresponding Egyptian hiero- glyphics and as illustrated by Pl. XVI, No. 37, of the present work. 4It closely approaches the original picture explained by a Babylonian scribe on the famous fr. from Kuyunjik, col. III, 6 (Trans. Soc. Bibl. Arch. VI, p. 455). 5 Cf. also the same sign on the very ancient monument preceding Urukagina’s time (De Sarzee, J. c., Pl. 1bis b., col. IV, 1). 6 As I have to dispose of more urgent matters at present, some years may still pass before its publication. TOnly his barrel cylinder in clay exhibits traces of the older form for mw, as shown above. 8 Nobody can object that a few characters in these Nippur inscriptions Seem to show the beginning of wedge- writing and that a few others Seem to have a later form. Lugalzaggisi presented c. 10) large inscribed vases, all apparently bearing the same long inscription here published, to Inlil of Nippur. Every stonecutter available was employed. Several of them understood but little of writing, and consequently some very ridiculous forms were produced. Cf., e. g., col. II, 16 (second variant), dug-a (sic/), 29 (second variant) da, 39 (variants) aga, 42 gur, 44 (fourth variant) ganam, 45 shig, and others. In order to understand the enormous difficulties which I had to over- come in restoring this text, Assyriologists will bear this fact in mind. 256 OLD BABYLONIAN INSCRIPTIONS ‘inscriptions very strongly. In the inscriptions of Edingiranagin, or Edingiranatum,' the grandson of Ur-Nina, a city, generally transliterated as Js-ban™, plays a very important réle. In fact the annihilation of the power of this city in S. Babylonia is the one prominent feature which characterizes his government, and to which (in connection with Erech, Ur and some other cities) the king refers again and again.” The most interesting object yet found in Tello, the so-called stele of vultures, was doubtless set ‘up by this sovereign in commemoration of his great victory over *’BAN“? How- ever this may be, so much is certain that at some time previous to Edingiranagin, a foreign power whose centre was %*’BAN"”, had succeeded in invading and conquering a large portion, if not the whole, of Babylonia, Erech and Ur included. The same eity of ”’ BAN” is also mentioned in the long Nippur text No. 87, and here again it occurs in connection with Erech and Ur (and Larsam). We learn at the same time from this very important historical document that Lugalzaggisi, son of a certain Ukush “natesi of ”’ BAN”! (col. I, 3,9, 10) had conquered all Babylonia and established an empire extending from the Persian Gulf to the Mediterranean Sea, in size there- fore not inferior to that founded much later by Sargon I. This first “king of the world” (lugal. kalama, col. 1, 4, 36-41, col. III, 4) of whom Babylonian documents give us information, selected Hrech as his capital, and by his great achievements raised ”' BAN", his native city, “to great power” (@ mag mu-um-gur, col. II, 41f.). The two documents, Nippur, No. 87, and the stele of vultures from Tello, belong closely together and supplement each other, the one giving a réswmé of the rise and height of the power and influence of “” BAN”, the other illustrating its downfall. The former must therefore antedate the monument of Edingiranagin. As doubtless some time elapsed between the rise and downfall of this foreign power; as, moreover, Shirpurla is not mentioned in Lugalzaggisi’s inscription, apparently because it did not as yet exercise any political influence ;° and finally as paleeographically this inscription from Nippur shows more traces of originality than the texts of Urukagina and Ur-Nina, as 1Tn view of De Sarzee, J. ¢., Pl. 31, No. 2, col. IIL, 5 (#-dingira-na-tum-ma = ‘‘ Brought into the house of his god”’ (by his parents after his birth). 2 Cf. De Sarzee, U. c., Pl. 8, Fragm. A, col. I, 5, 8, col. II, 4, 13, col. III, 5; Pl. 4, Fragm. A, col. II, 2, 11; Fragm. B, col. III, 8, col. V, 4; Pl. 31, No. 2, col. I, 6. * For details cf. Heuzey’s explanation of the figurative representations in his work, Les Origines Orientales, pp. 49-84, and in De Sarzec, l. ¢., pp. 174-184. LI agree with this scholar that the people whose defeat is illustrated on this monument belong to the city (and country) of giskBAN*i (De Sarzec, J. c., pp. 182). ‘This was the original reading of ]. 10; the traces preserved on two fragments establish my text restoration of this line beyond doubt. °The fragment of an inscribed object, apparently dedicated by a king of giskBAN” to Ningirsu, was found in Tello (De Sarzee, 1. ¢., Pl. 5, No. 8, and p. 119). From the character used for ‘‘ king’’ I draw the conclusion (with Heuzey) that the object belongs to a somewhat later period. Apparently gishBAN*i played a second important réle in the Babylonian history. CHIEFLY FROM NIPPUR. 257 stated above, we are justified in placing Lugalzaggisi before these two rulers of Shir- purla and in regarding most of the inscriptions published as Nos. 86-112 as older than the earliest royal inscriptions from Tello.1 At any rate, they are not later than these. A question of fundamental importance for our correct conception of the earliest phase of Babylonian history has been repeatedly discussed within the last ten years: In which relation did Sargon I (and Naram-Sin) stand to the early kings of Tello? Did he antedate or succeed them? Winckler’ and Maspero® expressed themselves decidedly in favor of the former view,* while Hommel,’ Heuzey ° and myself (Part I, p. 19),’ with more or less emphasis placed Sargon I and his son after Ur-Nina and Edingiranagin I will now briefly give the definite proof of the validity of our theory. 1. The results of the exploration of the lowest strata of Ekur will have convinced. us that Babylonian civilization had a history antedating the kingdom of Sargon I by ‘several thousand years. This pre-Sargonic period must have had a system of writing ; for the earliest texts at our disposal, however closely approaching the original picture in a number of cases, presuppose an earlier stage of writing, such as is testified to have existed in Babylonia by the monument “ Blau”* and by the famous fragments from ‘Kuyunjik” Pieces of inscribed objects unearthed below the Sargon level prove posi- tively that writing existed in Nippur long before Sargon I. It seems, therefore, at the ‘very outset, impossible to believe that not one document antedating the highly devel- oped style of writing in Sargon’s monuments should have been excavated in Nuffar or Tello. In fact, it would be altogether unreasonable to regard the inscriptions of Sargon and Naram-Sin as the first written records of the ancient Babylonian civili- -‘Zation. 2. Everybody who has studied the earliest inscriptions of Babylonia from their originals, and has devoted that special pains to all the details of paleeography, which 1The little fragment No. 107 cannot be referred to the time of Entemena, the only other ruler of Tello who, ‘according to our present knowledge, presented an inscribed vase to Inlil. Perhaps it is the first indication of the rising of Shirpurla in the South and of the extending of its sphere of influence northward at the expense of gishBANK, 2 Untersuchungen, p. 43; Geschichte, pp. 40f. (but cf. on the other side p. 42!); Altorientalische Forschungen III, ‘pp. 236ff. : 3TIn Recueil XV, pp. 65f.; The Dawn of Civilization, p. 605, note 3 (end). *Recently adopted by Rogers, Outlines of the History of Early Babylonia, Leipzig, 1895, p. 11, note 1 [but given up again after hearing my address, Contributions to the History of Sargon Iand His Predecessors, before the Oriental Club of Philadelphia]. 5 Zeitschrift fur Keilschriftforschung Il, p. 182 ; Geschichte Babyloniens und Assyriens, p. 291. 6 Cf., ¢. g., Les Origines Orientales, pp. 50, 84; Revue a’ Assyrtologie IL, pp. 54, 57. 7Cf. also Recent Research in Bible Lands, pp. 66f. 8 Called so for the sake of brevity. Cf. above, p. 249, note 4. ® Published by Houghton in Trans. Suc. Bibl. Arch., p. 454, and reproduced in several other works. 258 OLD BABYLONIAN INSCRIPTIONS I have a right to expect from those who criticise my statements on this subject, must necessarily come to the conclusion that a much longer period of development lies be- tween Lugalzaggisi, Urukagina, Ur-Nina and Edingiranagin, on the one hand, and Sargon and Naram-Sin, on the other, than between the latter and Ur-Ba’u Gudea, Ur-Gur, etc. It is surely remarkable that Monsieur Heuzey' and myself, who have devoted years of constant study to the paleeography of the earliest original inscriptions of Babylonia, quite independently of each other, have reached exactly the same conclusions. It is out of regard for the view of those who do not accept Nabonidos’ 3200 years as correct, that on paleeographic evidence alone I assign to Lugalzaggisi the minimal date of 4000 B.C. My own personal conviction, however, is that he can- not have lived later than 4500 B.C. 3. That my determination of the age of Lugalzaggisi is not too high is proved by the discovery of an uninscribed vase of precisely the same material and character- istic shape’ as most of the vases which bear Lugalzaggisi’s inscription. It was found 1.54 m. below the pavement of Naram-Sin, and must therefore considerably antedate the rule of the latter. 4, From paleographic and other reasons, I came to the conclusion above, that the inscriptions of Lugalzaggisi and of the other kings, patesis, ete., from Nippur grouped together with them, are surely older than Edingiranagin. IJleuzey, on the basis of other arguments, had inferred that the stele of vultures and the reliefs of Ur- Nina are “surely older than Naram-Sin.” Hence it would follow, that if Heuzey’s judgment of the age of these specimens of art is correct, also the monuments of Lu- galzaggisi, ete., antedate Naram-Sin. I am now in the position to prove the correct- ness of Heuzey’s view beyond question. Since a specimen of the workmanship of the artists at Naram-Sin’s time was recently discovered (ef. Pl. XXII, No. 64), showing exactly the same high degree of execution as the script on his monuments, every Assyri- ologist is enabled to judge for himself as to the value of Ileuzey’s judgment. There are, however, a few fragments of a relief in clay lately discovered in Nippur, which must be regarded as the strongest evidence in favor of the French scholar’s determination. While Heuzey declared Ur-Nina’s and Edingiranagin’s reliefs to be of greater anti- 1Tt is needless to quote passages from Mr. Heuzey’s works in addition to those given on p. 257, note 6. In connec- tion with his discussion of the age of the stele of vultures he makes the emphatic statement, ‘“‘le type linéaire de l’écriture est assurément plus ancien que celui des inscriptions de Naram-Sin,ete.’’ (cf. Les Origines Orientales, p. 50). ?Haynes reported on this vase, August 10, 1895, expressing the hope that I might be able to use it in support of my theory as to the age of most of the other ancient vase fragments from Nippur. He found it covered with earth and black ashes. It consists of white calcite stalagmite and has a very characteristic shape never found at a later period in Nippur again. In general this class of vases resembles a flower-pot, the diameter at the top being larger than that at the bottom, while the walls frequently recede a little at the middle. The size of the above-mentioned vase is : h. 26.5 ; d. at the top, 18; at the bottom, 14.8; at the middle, 13.8 cm. , CHIEFLY FROM NIPPUR. 259 quity than Naraém-Sin’s monuments, he characterized the relief which opens the splen- did series of De Sarzee’s finds (Pl. I, No. 1), and has several points of contact with the art exhibited in the stele of vultures, as “plus primitif, méme que celui de la grossicre tablette du roi Our-Nina” [De Sarzee, l. c., Pl. 1, No. 2], and as “une ceuvre @une antiquité prodigicuse, un monument des plus précieux, que nous devons le placer avee respect tout a fait en téte des séries orientales, comme le plus ancien example connu de la sculpture chaldéenne.” These words of a true master of his subject have found a splendid confirmation in the clay reliefs of Nippur just referred to, which in their whole conception and execution show a striking resemblance to the oldest spe- cimen of art recovered from Tello. They were found 7-7.70 m. below the level of Naram-Sin’s pavement, and within about 1.50 m. of the lowest trace of Babylonian civilization.' ‘Truly the genius and critical penetration of Heuzey could not have won a more brilliant victory. d. In connection with my examination of the pre-Sargonic strata of Ekur, I twice called attention to the fact that baked bricks found below Naram-Sin’s pavement are plano-convex in form.” I might have added that no other form of baked brick has so far been discovered anywhere in the lowest strata of Nippur, and that these bricks as a rule bear a simple thumb mark upon their convex side. The form of these baked bricks, until the contrary has been proved, must therefore be regarded as a character- istic feature of all structures previous to the time of Sargon I and Naram-Sin. It is quite in accordance with this view that the only inscribed bricks of Tello which show this peculiar form, bear the legend of Ur-Nina, whom on other evidence I placed before Sargon and Naraim-Sin. 6. We draw a final and conclusive argument from a door-socket of Sargon him- self. In Part I, Pl. 14, Nos. 23-25, I published three brief legends of a king whom, influenced by Pinches’s reading (Garde), I read Gande (pp. 28 ff.), and whom I regarded as identical with Gandash, the founder of the Cassite dynasty. All that I brought forward in favor of this identity I herewith withdraw ; when I wrote those 1Cf. above, p. 240, note 2. They will be published in Series B of the expedition work edited by myself. 2 The bricks of the ancient curb around the altar, p. 238, and the bricks of the ancient arch, p. 24). In his report of Oct. 26, 1895, Haynes refers to the discovery of a terra-cotta floor with a rim a little below the pavement of Naram- Sin. He regards it as a combination of bath and closet, ‘‘ proving that the present customs and methods of preparing the body for worship, as practiced by Moslems [in the immediate neighborhood of their mosques], is of very great anti- quity. The drainage from this floor was conducted into a large vertical tile drain, which is 2 m. long and has an ’ average diameter of 85 cm.”’ This tile drain is ‘‘supported by a double course of bricks, plano-convex in form, with finger marks on the convex side.’’ Fora specimen of Ur-Nina’s bricks cf. De Sarzec, J. c., Pl. 31, No.1. Specimens of this class of Nippur biicks were given by Petersin The American Archeological Journal X, p. 34 (two drawings from the hand of the late Mr. Mayer, + 20 Dec., 1894, in Bagdad). The peculiar shape of these bricks in the arch is scarcely distinguishable on Pl. XXVIII of the present work. 260 OLD BABYLONIAN INSCRIPTIONS. pages, I was still somewhat influenced by the current view of Assyriologists, that later kings occasionally imitated older patterns in their script. Since then I have completely shaken off this old theory as utterly untenable when contrasted with all the known facts of Babylonian paleography. The observation, however, which I made on p. 29, note 2, that the characters represent the peculiarities of Ur-Nin4’s inscriptions was entirely correct. Since then a large number of vase fragments have been exca- vated, by which I was enabled to confirm and strengthen my previous judgment based upon the study of a few squeezes of badly effaced inscriptions and to analyze the pal- sographic peculiarities of this whole class of ancient texts completely. I arrived at once at the result that the three legends published on Pl. 14 were written by Lugal- kigub-nidudu, “lord of Hrech, king of Ur,” who left us No. 86. Among other gifts, such as vases, dishes, etc.,' this sovereign presented a number of unhewn diorite, ealcite, stalagmite and other blocks” to the temple as raw material for fature use.*® At the time of Bur-Sin II several of these blocks, of which one is published on Pl. X VIT, were still unused." They had been handed down from a hoary antiquity and scrupu- lously preserved for c. 1500-2000 years in the temple archive. Bur-Sin IT selected a diorite block from among them, left the few words of its donor respectfully on its side,’ turned it into a door-socket, wrote his own inscription on its polished surface and pre- sented it in this new form to the temple. But something similar happened many hun- dred years before. According to Part I, p. 29, section 1,° the same rude inscription is scratched upon the back side of a door-socket of Sargon I. From the analogous case just treated it follows that Lugal kigub-nidudu must have lived even before Sargon J, and consequently that all other inscriptions which have the same paleeographie peculi- arities as his own can only be classified as pre-Sargonic. 1Cf. Pl. XVIIL, 40-48. 2Cf. Part I, p. 29. $ These blocks received therefore only a kind of registering mark scratched merely upon their surface (Dingir En- lil(-la) Lugal-ki-gub-ni-dudu (ne) a.mu-na-shub, ‘‘To Inlil L. presented (this’? =7e)). The inscription on the block, Pl. XVII, No. 39, had originally 8 li. according to the traces left. On the diorite blocks these inscriptions are well preserved; on the calcite blocks however, whose surface corroded and crumbled in the course of six millenniums, they have suffered considerably. Cf. on the whole question of presenting stones as raw material to the temple, Hilprecht in Z. A. VIII, pp. 190 ff. * As shown above. / ° Cf. The curses on the statue B of Gudea, col. VII, 59 ff., on the door-sockets of Sargon, Pl. 1, 12 ff., Pl. 2, 13 ff, on the lapis lazuli block of Kadashman-Turgu, Pl. 24, pp. 14-20. In the latter case the lapis lazuli was likewise pre- sented as raw material to be used in the interest of the temple. But the inscription—this was the intention of the donor—was to be preserved (a thin piece of lapis lazuli being cut off, cf. Pl. XI, No. 25) in remembrance of the gift. ®Cf. Part I, ‘‘Table of Contents,”’ p. 47. bo or r— CHIEFLY FROM NIPPUR. CONTENTS AND HISTORICAL RESULTS. In the briefest possible way I will indicate the general results which I draw from a combined study of the most ancient Nippur and Tello inscriptions. With the very scanty material at my disposal this sketch can only be tentative in many points. For every statement, however, which I shall make, I have my decided reasons, which will be found in other places.’ At the earliest period of history which inscriptions reveal to us, Babylonia has a high civilization and is known under the name of Heng, “land of the canals and reeds,”” which includes South and Middle Babylonia and possibly a part of the North. Its first ruler of whom we know is “ Hn-shagsag-ana, lord of Kengi.”’ Whether he was of foreign origin or the shaykh of a smaller Babylonian “city” which extended its influence or the regular descendant of the royal family of one of the larger cities, can- not be decided. It is therefore impossible to say whether he belonged to the Sumerian or Semitic race, or traced his origin to both. That the Semites were already in the country results, aside from other considerations,‘ from the fact that the human figures on the stele of Ur-Enlil, which belongs to about the same period,’ show the characteristic ‘In Assyriaca, part II, in Z. A., and in response to a repeated invitation from the President and Secretary of the Philosophical Society of Great Britian, in the Z/ansactions of the latter society, where I expect to give a more complete sketch of the political and social conditions of ancient Babylonia. 2Cf. No. 90, 4 (also No. 87, col. II, 21) and above p. 252, note 9. 3 His inscriptions (Nos. 90-92) have the oldest form of mw, have older forms for say and show other characteristic features of high antiquity. His name signifies ‘‘lord is the king of heaven.”’ *Cf. for the present only the important argument drawn from Lugalziggisi’s inscription No. 87, col. IIL, 35. Here we have the same writing DA-UR, which from the inscriptions of Nebuchadrezzar IL and other latest Babylonian kings, is known to be a Semiticism for daru. Cf. Delitzsch, Assyrisches Handworterbuch, p. 218. °It has the most ancient forms for dam and mw and shows a very characteristic feature of the oldest period of writing by contracting the name of Min-din dug(-ga), or Ba’u (cf. above p. 252) into a monogram. The primitive style of art, and such details as the headdress of the god, the short garment of the two persons following the sheep and goat, the nakedness of Ur-Ealil, the fact that his figure and the other two have their hair shaved off, corrob- orate my determination of the age of this monument. On the other hand, this stele and No. 38 of the same plate, which doubtless belongs to the same age, show us a real Old Babylonian master, who produced a beautiful ensemble with a few simple lines, and knew how to breathe life into his very realistic but very graceful figures. Cf. the great skill he exhibits in his drawing of the graceful outlines of a gazel, and his remarkable knowledge of animal locomo- tion ! The two animals in No. 37 ‘represent very characteristically two species, the near one a goat and the far one a sheep. The goat shows more characteristics of the wild species of Eastern Persia and Afghanistan than of the Per- sian, and so may be a domestic hybrid between the two (@. ¢., Cupra fulconerii and Cupra @egagrus). The sheep is probably also derived from Eastern Persia and is perhaps the ‘ urial’ Ovis vigned, which is an ally of the domestic sheep. It has resemblance also to the Armenian wild sheep Ovis gmelinii, but the rugosity of the horns is too great, and the lines are too vertical’ (communication from my colleague, Dr. Edward D. Cope, Professor of Zodlogy and Comparative Anatomy, who kindly examined the monument). A. P. 8.— VOL. XVII. 2 H. 262 OLD BABYLONIAN INSCRIPTIONS features of a mixed race.' The capital of this early kingdom is likewise unknown.’ In all probability it was Erech.’ The religious centre of Kengi was the sanctuary of Inlil at Nippur.* It stood under the especial care of every ruler who claimed supreme authority over the country, and who called himself pates? gal Inlil,’ to define his posi- tion as being obtained by divine authority. The chief local administrator of the tem- ple in Nippur seems to have had the title damkar gal.° This I infer from my analysis of the meaning of damkar and from the inscriptions of Nos. 94 and 95 in connection with No. 96, where a certain Aba-Inlil (= Aishit-Bél) who has the title of damkar, presents a vase to Ninlil “ for the life of Ur-Inlil, patesi of Nippur.’” Ur’ and Larsam’ and doubtless other places whose names are not yet known from inscriptions, were prominent cities in this early Babylonian kingdom. 'They had their own sanctuaries, which stood under the control of a pates?. ‘This title characterizes its bearer, according to his religious position, as sovereign lord of a temple and chief servant of the god worshiped in it. The fact that a patesi, in addition, often occupied a political position as king or governor, does not interfere with this view. He is first of all the highest official of his god, representing him in his dealings with his subjects; in other words, 1 Prof. Cope wrote me on this subject : ‘‘The shortness of the jaws however is certainly not a Semitic character in human faces, and this character renders the physiognomy very peculiar. The hooked nose and large eyes on the con- trary are Semitic. Asa result I should say the figures represent an Aryan race with some Semitic tendencies. The identification of such a race is of much interest [indeed it is of vital importance for the whole Sumerian question ! —H.]. The people evidently have no Mongolian tendencies.”’ *It may have stood in No. 90, 5, Jugal. ..., which is only preserved in part. The traces do not point to the ideo- gram of Unug, more to kalama. 3 Cf. Nos. 86, 4-14; also the fact that Erech is the capital of Lugal-kigub-nidudu and Lugalzaggisi and is promi- nently mentioned in Edingiranagin’s inscriptions. Cf. also Hommel, Geschichte, p. 206, and especially p. 300, observe the important position which Erech holds in the titles of the kings of the dynasty of Isin en (shega) Unugaki [V. B. Winckler’s reading of Pait I, No. 26, 3, as Stn-ga-mil, is an absolute paleeographic impossibility. If anything, the reading of this line as Unugki-ga-ge is sure beyond question (against Winckler, Altorientalische Forschungen III, p. 274)]. *Cf. above, p. 236, and among other points, especially No. 87, col. I, 36-41. ® Cf. No. 87, col. I. Literally ‘‘ate’’ (akalw) or “was filled with ”’ (shuznunuw). ® The variant is a peculiar form of ga (not =7zgz), cf. col. ILL, 21, 23 and variants. TNo. 87, col. I, 1. DingirEn-lil 2. lugal kur kur-ra 3. Lugal-zag-gi si 4. lugal Unugki -ga 5. lugal kalam-ma 6. shib An-na 7. galu mag 8. UngirNiduba 9. dumu U-kush 10. [pa-tle-s¢ gshBA NK 11. galu may 12. dingir Niduba-ka 13. gt 2% bar-ra 14, dingir Lugal-kur.kur-ra 15. pa te-si gal 16. dingir Hn-lil 17. gish-PI-SHU-sum-ma 18. dingir ENKI 19. mu-pad- da 20. dingir Jiu 21. luy may 22. dingir Hn-2u 2B. ne-gish 24, dingir Utu 25. u%-a Uingir Ninna 26. dumu tu-da 27. dingir Ni-daba 28. ga 2i ku-a 29. ding Nin-har. sag 80, galu dingirUmu sanga Unugt-ga 31. sag eyi-d 82. dingir Nin-a-gid-ya-du 33. nin Unug*i -ga-ka 34. iti (2) may 35. dingir-ri-ne-ra. CHIEFLY FROM NIPPUR. 267 Fast, and yet his very name had been forgotten by later generations. He lived long before Sargon I founded his famous empire, and he called a kingdom his own which in no way was inferior to that of his well-known successor, extending from the Persian Gulf to the shores of the Mediterranean. I quote the king’s own poetical language: “When Inlil, lord of the lands, invested Lugalzaggisi with the kingdom of the world and granted him success before the world, when he filled the lands with his renown (power) (and) subdued (the country) from the rise of the sun to the setting of the sun—at that time he straightened his path from the lower sea of the Tigris and Kuphrates to the upper sea and granted him the dominion of everything (?) from the rise of the sun to the setting of the sun and caused the countries to rest (dwell) in peace.” * It becomes evident from this passage, in which Lugalzaggisi declares him- self to have been invested with the kingdom of the world by Inlil of Nippur, “lord of the lands,” that only Nippur can have been the ancient seat of the sharrit kibrat arbw’im, which manifestly is but the later Semitic rendering of the ancient Sumerian nam-lugal khalama. Ihave examined all the passages in the fresh light of this text and find that Nippur fulfills by far better the required conditions than Kutha or any other city which has been proposed in Northern Babylonia. But, be it remembered, to the early kings of Babylonia this title meant more than a mere possession of the city whose god claimed the right of granting the sharrit kibrat arbwim. Down to the time of Hammurabi only those* laid claim to this significant title who really owned territory far beyond the north and south of Babylonia, who, in the Babylonian sense of the word, had conquered a quasz worldwide dominion, defined by the four natural boundaries (Part I, p. 25). The later Babylonian and Assyrian inscriptions are of value for the determination of the meaning of this title at their own time, but they have little importance for the question as to its origin and earliest localization, if the title must be localized at all hazards. According to the manner of usurpers,’ Lugalzaggisi retained Hrech, the old metropolis of the country, as his own new capital of this first great Oriental state, of which Kengi became now the chief province. Babylonia, as a whole,‘ had no fault 1Col. I, 36. Ud dingirEn-lil 37. lugal kur-kur-ra 38. Lugal-zag-gi-st 39. nam-lugal 40. kalam-ma 41. ma-na-sum-ma-a 42. igt kalam-ma-ge 43. si ma-na-di-w 44. kur-kur(a)ne na 45. ma-ni-sig-ga-a 46. Utu e(a)-ta. Col. Il, 1. Utw shu(a)- shu 2. gu ma-na-gar-ra-a@ 3. uda-ba 4. a-ab-ba 5. sig-ta-ta 6. Idigna 7. Buranunu(without determ.)-bi(= “‘and’’) 8. a- ab-ba 9. igi nim-ma-shu10. gira-bi 11. st-mi-na-di 12. Utu e(a)-ta 13. Utu shu(a)-shu 14. [dingirH]n-Ul-té 15... .. nin 16... .. . mu-ni-dug 17. kur kur(a) u sal-la 18. mu-da-na. 2 Of Dungi we know too little to call him an exception. Of the kings of the second dynasty of Ur, who assumed the proud title, we know now from Pls. 55 and 58 (cf. above, p. 246 and note 4) that they had made conquests as far as Syria and Elam. * Well stated by Winckler, Altorientalische Forschungen III, p. 284. *Cf. col. II, 19. kalam-ma 20. a-ul-la mu-da-ga (= shakanu) 21. bar-bar Ki-en-gi 22. pa-te-si kur kur-ra, ete., ete. 268 OLD BABYLONIAN INSCRIPTIONS to find with this new and powerful régime. The Sumerian civilization was directed into new channels and prevented from stagnation; the ancient cults between the lower Tigris and Euphrates began to revive and its temples to shine in new splendor. Hrech, Ur,’ Larsa* and Nippur’ received equal attention from their devoted patesi. But first of all, “’BAN" itself, the native city of the great conqueror, was raised by his energy and glory to a position of unheard-of influence and political power. Lugalzaggisi stands out from the dawn of Babylonian history as a giant who deserves our full admiration for the work he accomplished. He did not appear unexpectedly on the scene of his activity. We had been prepared for the collapse of the ancient monarchy on the Persian Gulf, with its long but unknown history, by the preceding invasions and victories of the Northern hordes to which he belonged. And yet when suddenly this great empire of Lugalzaggisi stands before our eyes as a fat accompli, we can scarcely conceive, whence it came and how it arose. There is no doubt in my mind that Lugalzaggisi’s achievements in Babylonia represent the first signal success of the invading Semites from the North. On the previous pages we have seen how these hordes were pushing gradually southward. After for a number of years they had concentrated their attacks upon the border forti- fications of Northern Babylonia and had established a military station and kingdom in Kish, it was but a question of time when the whole country in the South had to suc- cumb to their power. The oldest written monuments of Babylonia do not designate these enemies by any single definite name: they are the hordes of the city of “’BAN® and Kish combined, apparently but two centres of the same powerful people which was roaming over the fertile steppes of Mesopotamia, and whose chief stronghold doubtless was “"BAN". What ancient city, then, is this "BAN"? That we have not to place it “in Susian territory,” as Maspero‘ is tempted to do, is beyond question. The ideogram for lugal on an inscribed object of Tello and presented by a king of *"BAN™ (De Sarzec, l. c., Pl. 5, No. 3), points with necessity to the north for the location of our city. As this peculiar form of the character for lugal so far has only been found in such cuneiform inscriptions as contain Semitic words written phoneti- cally, or in other texts which are written ideographically, but, on the basis of strong arguments’ must be read as Semitic, we are forced to the conclusion that this charac- 1Col. II, 30-82. Urumki-ma guda-gim sag-ana-shu mu-um-gur, ‘Ur like a steer he raised to the top of heaven.” 2 Col. IL, 83-37. Larsamki ur ki-ag dingirUtu-ge a-ne-gul-la mu-da-ga. For giskBANhi cf. ibidem, 38-42. ’ As becomes evident from his titles and from the extraordinary number of vases presented io Inlil. ‘The Dawn of Civilization, p. 608. Cf. also Heuzey in De Sarzee, 1. c., p. 182. 5 Cf. for the present above, p. 263, note 1. More on this subject and on ‘‘the Semitic influence in early cuneiform writing in general in another place. My above statement is the result of a complete and exhaustive examination of all the published cuneiform material in which the peculiar form of Jugal occurs. CHIEFLY FROM NIPPUR. ; 269 ) ter, while doubtless derived from the well-known Sumerian form, was invented and employed by a Semitic nation. Furthermore, I-call attention to the important fact that Lugalzageisi, who was surely a Semite,' shows his nationality in various ways, such as the use of certain phrases, which look very suspicious in an ancient Sumerian inscription,’ and especially in his use of the ideogram da-ur, doubtless of Semitic origin (= déri), for “eternal.”* There is only one ancient place in Northern Meso- potamia which could have been rendered as “the city of the bow” ideographically by the Sumerians, namely Harran, with which %*"BAN*“ is doubtless identical. For according to Arabic writers, especially Albirdnz (ed. Sachau, p. 204), the ground-plot of Harran resembled that of the moon (7. ¢., the crescent or half-moon), and Sachau, who gaye us the first accurate sketch of this city, finds it very natural that “ Arabic writers could conceive the idea of comparing it with the form of the half-moon.”? Excellent, however, as this Arabic description is, and valuable as it proves for our final location of ”"BAN™, the ancient Babylonian ideographie rendering as “ city of the bow ” was a more faithful description of the peculiar way in which Harran was built than any other, as everybody can easily convince himself by throwing a glance upon Sachau’s plan in his Rezse in Syrien und Mesopotamien. This correct solution of a vexed problem becomes of fundamental importance for our whole conception of the history of the ancient Hast. First of all, I have furnished a better basis for Winckler’s ingenious theory of the original seat of the sharrit kishshati. All that could be gath- ered from later historical sources, beginning with the end of the second millennium before Christ, Winckler brought together to formulate a view which never found much favor with Assyriologists and historians.’ I opposed it myself’ on the ground that his reasons proved nothing for the ancient time, because Harran was never mentioned in a text before the period just stated, and that in view of the total absence of a single 1Tf he did not adopt a Sumerian name when ascending the throne of Kengi and of the ‘‘kingdom of the world,’’ which is very probable, the name of the king must be read something like Shurru-mali-emuki-kent (emuku is masce. and fem. in the singular). But the name cannot be regarded as the prototype of Sargon I (= Sharru-kénw), because, aside from other reasons, this kind of abbreviation of a fuller name is without parallel in the history of Assyrian proper names. They are abbreviated at the beginning or end, but not in the middle. Cassite names, etc., are foreign names. 2Of., e.g., ‘from the lower sea of the Tigris and Euphrates to the upper sea,”’ ‘‘ from the rising of the sun to the setting of the sun’”’ and others, which remind us forcibly of the phraseology of the latest Assyrian monarchs. 3 Co]. III, 36. da-ur ye-me, ‘“he may pronounce (speak) forever !’’ 4Cf. also Mez, Geschichte der Stadt Harran in Mesopotamien, p. 9. The remark of the Arabic writer is therefore more than a ‘‘ Treppenwitz,”’ and is of great historical importance, showing us that not only the ancient Babylonians but other peoples were struck by the remarkable form in which Harran was built. 5Sachau, Reise in Syrien und Mesopotamien, p. 223. 6 Cf. especially Winckler, Altorientalische Forschungen I, pp. T5ff.; III, pp. 201 ff. 7Part I, pp. 23 f. I was supported in this, e. g., by Jensenin Z. A. VIII, pp. 228 ff A, B, S—yOin, Sadi, Ait 270 OLD BABYLONIAN INSCRIPTIONS reference to this city in our whole ancient literature previous to 1500 B. C., we could not speak of it as the seat of a kingdom until we first proved that the city really ex- isted. From the fact that (1) Aish and Kesh (shatu) did not only sound alike but were eyen used interchangeably in the inscriptions,’ (2) that many other ancient Babylonian cities (cf. Shirpurla)’ are frequently written without a determinative, (3) that the city of Kish played a very important réle in the inscriptions of Edingirana- gin,’ (4) that all the ancient empires arose from city kingdoms, and from several other considerations,‘ I inferred that shar KISH meant originally “king of Kish,” a com- bination which Winckler himself regarded “naheliegend.’”” But notwithstanding the great importance which must be attached to the kingdom of Kish in connection with the final overthrow of the ancient empire of Keng, Kish was not the principal leader in this whole conquest, but was controlled by a greater power in the North, Harran, as I have shown above. Having therefore demonstrated the existence of the city of Harran at the threshold of the fifth and fourth pre-Christian millenniums, which Winckler failed to do, although Hdingiranagin’s inscriptions, which necessarily formed the starting point of my operations, had been at his disposal for some time, and hay- ing furthermore indicated the powerful position which Larran must have occupied as the great Semitic centre of the ancient Orient, I am now prepared to accept Winckler’s theory of the original seat of the sharrit kishshati without reserve. I regard the title as the Assyrian equivalent of the Sumerian nam-lugal kalama. In view of the lead- ing part that Harran had taken in the establishment of the first “kingdom of the world ” under Lugalzaggisi, Harran became the seat of the Semitic sharrit kishshate just as Nippur was the centre of the Sumerian nam-lugal kalama. When after many vicissitudes under Sargon I and Naram-Sin finally the northern half of ancient Kengi, including Nippur, was definitely occupied by a Semitic population, which spoke and wrote its own language, the old Sumerian title nam-lugal kalama, which carried the same meaning for the inhabitants of Babylonia as sharrit kishshati did for 1Cf. Winckler, J. c., pp. 144 f. 2 Tn the inscriptions of Ur-Nina written without x7. ’ Not only in his stele of vultures, but also in the inscription unearthed in London (Proce. Soc. Bibl. Arch., Nov., 1890). Hommel was of the opinion (Die /dentitat der altesten babylonischen und agyptischen Gottergenealogie, p. 242), that the passage in the Jatter text escaped my attention. I simply had no use for it: (1) lugal Kish an ki is some- thing entirely diflerent from lwgal an-ub-da tab-tab-ba or lugal KISH; for if it was possible to say so in Sumerian, it could only mean ‘“‘king of the whole heaven and earth,”’ which the king of course did not want to say. (2) The text does not offer this at all, but must be translated lugal Kishki-bi-na-dib-bi, ‘and the king of Kish,’ in other words dvis copula = “and,” connecting Kish*i with what stood before. Cf. in the present work, PJ. 87, col. II, 7 (‘‘and ” the Euphrates). *Cf. Part I, pp. 23 f 5 Altorientalische Forschungen Il, p. 145, note 1. CUIEFLY FROM NIPPUR. 271 the Semites of Northern Mesopotamia, disappeared and was translated into the Sem- itic sharrit kibrat arbwim. The later Sumerian nam-lugal “ub-da-tab-tab-ba is nothing but a translation from the Semitic title back into the sacred Sumerian lan- guage by Semitic scribes of the third millennium B. C. Not long after Lugalzaggisi’s death a reaction seems to have set in. Sugir gen- erally transliterated as Girsu, which Urukagina or one of his predecessors raised from the obscurity of a provincial town to the leading position in the new kingdom of Shir- purla, must be regarded as the centre of a national Sumerian movement against the Semitic invaders. “The lord of Sugir,” Win-Sugir, became the principal god, and his emblem —the lion-headed eagle with outspread wings, occasionally appearing in connection with two lions, which are victoriously clutched in its powerful talons'—be- came the coat-of-arms of the city and characterizes best the spirit of independence which was fostered in its sanctuary. Urukagina’s successors, especially Ur-Nin4, devoted their time to building temples and fortifying the city of Shirpurla and, as faithful patesis, impressed the power and glory of their warlike deity upon their sub- jects. The cult of Nin-Sugir cannot be separated from the national uprising which started from his sanctuary. Edingiranagin at last felt strong enough to shake off the obnoxious yoke of the Semitic oppressors of Kish and Harran. ‘The decisive battle which was fought must have been very bloody. The Sumerians won it, and they cel- ebrated their victory, which restored a temporary power and influence over the greater part of Kengi to them, in the famous stele of vultures set up by Hdingiranagin. Erech and Ur played a prominent part in this national war. The former retained its place as the capital of the nam-en (of Kengi), but Ur seems to have furnished the new dynasty, as I infer from No. 86. Although No. 86 of my published texts belongs doubtless to the same general period as No. 87, a detailed examination of its palzeographic peculiarities leads me to place it somewhat later, and to regard it as about contemporary with the inscriptions of the kings of Shirpurla, especially with those of Edingiranagin. We learn from it the following: ‘“ When Inlil, the lord of the lands, announced life unto Lugal-kigub- nidudu, when he added lordship to kingdom, establishing Hrech as (the seat of ) the lordship (the empire) and Ur as (the seat of) the kingdom, Lugal-kigub-nidudu pre- sented this for the great and joyful lot (which he received) unto Inlil, his beloved 1Cf. Heuzey’s treatise Les Armoiries Chaldéennes. ? Five different legends have been found of this ruler: (1) A brief legend of three lines (cf. Pl. 14), (2) one of seven or eight lines (cf. Pl. XVIL No. 39), () one of nineteen lines, (t) an even larger one of c. thirty lines, (5) No. 88. Of the third class a fragment was excavated after the preparation of my plates, which contained the closing lines 17-19. The precise connection between the upper and lower portions on Pl. 37 cannot be given at present. 272 OLD BABYLONIAN INSCRIPTIONS CHIEFLY FROM NIPPUR. lord for his life.’ In Lugal-kigub-nidudu’ and his son (?) Lugal-kisal-si? we have therefore the first representatives of the first dynasty of Ur. Ur-Gur and Dungi, ete., who lived about 1000 years later, must hereafter be reckoned as members of the second dynasty of Ur.“ The relation of this dynasty to Edingiranagin is shrouded in absolute mystery. It is not impossible that its members ruled before him and were Semites who overthrew the dynasty of Lugalzaggisi. How long the restored Sumerian influence lasted we do not know. Apparently the Semites were soon again in possession of the whole country. The old name Keng? continued to live as an ideogram in the titles of kings; but the name of Shumer, by which Southern Babylonia was known to the later Semitic populations, was derived from the city of Sugir or Sungir,? which was the centre of the national uprising of the South against the foreign invaders from Kish and Harran. Sargon I finally restored what had been lost against Edingiranagin. In his person and work we see but a repetition of that which had happened under Lugalzaggisi centuries before. From the city of Agade,* which became the capital of the Sargonic empire, I derive Akkad, the name of Northern Babylonia. The names of Shumer and Akkad are therefore but the historical reflex of the final struggle between the Sumerian and Sem- itic races, and they were derived from the two cities which took the leading part TMs! 11. DingirEn-Ul. 2. lugal kur-kur(a)-ge. 3. Lugal-ki-gub-ni-du-dura 4. ud dingirEn-lil-li 5. gu-et ma-na-de-a 6. nam-en 7. nam-lugal(a)-da 8. ma-na-da-tab-ba-a 9. Unugki-ga 10. nam-en 11. mu-ag-ge 12. Urumki-ma 18. nam- lugal 14. mu-ag-ge 15. Lugal-ki-gub-ni-du-du ne 16. nam gal-yul-la-da 17. dingir Enlil lugal ki-alga-ni 18. nam-ti- la-ni-shu 19. a-mu-na-shubd]. The use of da = shu, ‘unto, for,’ in this text is interesting, cf. ]. 7and 1.16. We meet the same use in No. 111: 1. DingirNin-din-dug-ga 2. wma nin 8. dam 4. ff .... 3 f.e. SDugal-shir-ge 2. fe. nam-ti 1 f. e. dam- dumu-na-da a-mu-shub. * «The king finished the place’? = Sharru-manzazu-ushaklil. * Or Lugal-si-kisal, i. e., ‘The king is the builder of the terrace,’ Sharru shapik-kisallé, From the close connec- tion in which Lugal-kigub-nidudu, who left many fragments of vases in Nippur, stands with Lugal-si-kisal on Pl. 37, No. 86, 11 f. e—1, Iam inclined to regard them as father and son. Cf. also No. 89. *Cf. Hilprecht, Recent Research in Bible Lands, p. 67. ° Cf. alveady Amiaud in The Babylonian and Oriental Record I, pp. 120 ff. On the reading of Sugir instead of Girsw cf. also Hommel, Geschichte, pp. 290, 292, 296, etc., and Jensen, in Schrader’s K. B. III, part 1, pp. 11 f. (note). ® With George Smith, Amiaud, Hommel and others (against Lehmann, Shamashshumukin, p. 73). That Agade can go over into Akkad philologically, Ican prove from other examples. But even if this was not the case, the clear statement of George Smith (cf. Delitzsch, Paradies, p. 198) should be sufficient. I cannot admit the possibility of a original mistake on the part of George Smith. Master in reading cuneiform tablets as he was, he could not have made a blunder which would scarcely happen to a beginner in Assyriology. ‘That Akkad became finally identical with ‘‘the Babylonian empire in its political totality and unity,’’ was dem- onstrated by Lehmann, J. ¢., pp. 71 ff. TABLE OF CONTENTS AND DESCRIPTION OF OBJECTS. Part If, Plates 36-70 and XVI-XXX. ABBREVIATIONS. angul., angular; beginn., beginning; ¢., circa; ca., cast; C. B. M., Catalogue of the Babylonian Museum, University of Pennsylvania (prepared by the editor); ef., confer; col., column(s); Coll., Collection; d., diameter; Dyn., Dynasty; E., East(ern); f., following page; ff., following pages; f. e., from (the) end; follow., following; fr. or fragm., fragment(s), fragmentary; h., height; horizont., horizontal; ibid., ibidem; imser., inscription; ]. or li., line(s); m., meter; M. I. O., Musée Impérial Ottoman; N., North(ern); Nippur I, IL, ILI, etc., refers to the corresponding numbers on Plate XV; No., Number; Nos., Numbers; Obv., Obverse; omit., omitted; orig., original(ly) ; p., page; pp., pages; perpend., perpendicular; Pho., Photograph; PIl., Plate; re. or resp., respectively; Recueil, Recueil de travaux relatifs a la philologie et a l’archéologie égyptiennes et assyriennes, edited by G. Maspero; restor., restored; Rev., Reverse; S., South(ern); sq., squeeze; T., Temple of Bél; var., vari- ants; vol., volume; W., West(ern); Z., Ziqqurratu; Z. A., Zeitschrift fiir Assyriologie, edited by C. Bezold. Measurements are given in centimeters, length (height) X width X thickness. Whenever the object varies in size, the largest measurement is given. The numbers printed on the left, right and lower margins of Plates 36-42 refer to C. B. M. and denote the vase fragments used in restoring the cuneiform texts here published. If more than one fragment is quoted, they are arranged according to their relative importance. On fragments placed in parentheses, as a rule less than one or two complete cuneiform characters are preserved. Fragments originally belonging to the same vase are connected by + or-+x-, the former indicating that the breaks of fragments thus joined fit closely together, the latter that an unknown piece is wanting between them. I. AvroGrRArH REPRODUCTIONS. Prarn, Opxt! Dats. DESCRIPTION. 36 86 Lugal-kigub-nidudu. Fragm. of a large vase in serpentine, 20.5 X 9.45 X 2.8, orig. d. c. 25.4. Nippur III, beneath the rooms of T. on the 8. E. side of Z., a little above Ur-Ninib’s pavement in the same stratum as has pro- duced nearly all the fragments of the most ancient stone vases so far excavated in Nuffar (approximately therefore the same place as Pl.1, No.1). Inser. 15 (orig. at least 30) li. C. B. M. 9825. Portions of these 15 li. preserved on the follow. 21 other fragm. of vases in calcite stalagmite (from which the text had been restored before 9825 was found and examined): C. B. M. 9657 + 9607 + 9609 (ef. Pl. XVIII, Nos. 41-43), 9581 + 9648, 9608 + 9679 49591 (belonging to the same vase as 9900, cf. Pl. 387 and Pl. 274. PLATE. 37 38 43 TEXT. 86 87 87 87 87 87 88 89 90 91 OLD BABYLONIAN INSCRIPTIONS Date. Lugal-kigub-nidudu. Lugalzaggisi. Lugalzaggisi. Lugalzaggisi. Lugalzaggisi. Lugalzaggisi. Lugal-kigub-[nidudu]. Lugal-kisalsi. En-shagsag(?)-anna. En-shagsag(?) anna. DESCRIPTION. XVIII, No. 47), 9901, 9902, 9908, 9904 (cf. PJ. 37), 9905, 9632 (be- longing to the same vase as 9635 + 9620 + 9627 + 9606, ef. Pl. 37), 9605 (cf. Pl. XVIII, No. 44), 9599, 9633, 9680, 9708, 10001 (cf. Pl. XVIII, No. 48). Cf. also 9634 (cf. Pl. 37 and P]. XVIII, No. 46). The same inser. continued. On the scale of fr. 9325 restored from 16 fragm. of vases in white calcite stalagmite. Nippur III, approximately same place as Pl. 36. C. B. M. 10001 (ef. Pl. 36 and Pl. XVIII, No. 48), 9900 (cf. Pl. XVIII, No. 47, belonging to the same vase as 9608 + 9679 + 9591, cf. Pl. 36), 9904 (cf. Pl. 36), 9620 + 9627 + 9635 + 9606 (belonging to the same vase as 9632, cf. Pl. 36), 9604, 9630, 9631, 9917 (red banded), 9639, 9644. Cf. also 9634 (ef. Pl. 36 and Pl. XVIII, No. 46), 9607 (ef. Pl. 36 and Pl. XVIII, No. 41), 9613 (cf. Pl. XVIII, No. 40). Five fragm. of a vase in white calcite stalagmite (glued together), 16X138X1.9. Nippur III, approximately same place as Pl. 36, No. 86. Inser. 3 col., 13-+-17-+8=38li. C. B. M. 9914+ 9910 +9915+ 9913 + 9320. Cf. Pl. XIX, No. 49. On the basis of these five fragm. the complete text published on Plates 38-42 has been restored by the aid of the follow. 88 other fragm. belonging to 63 different vases: C. B. M. 8614, 8615, 9300, 9301, 9304, 9306, 9307 + x + 9668, 9308, 9309 + 9924 +. 9311 + 9316 + 9314 + 9916, 9312 (ef. Pl. XIX, No. 59), 9317, 9318 + 9645, 9583, 9584 + 9315, 9587, 9595, 9598, 9601 + 9305, 9602, 9611+ x-+ 9610 (cf. Pl. XIX, Nos. 50, 51), 9619, 9624, 9625, 9628 (cf. Pl. XIX, No. 53), 9638, 9642, 9646 + x + 9310, 9651 + 9911, 9654, 9656 + 9685 (ef. Pl. XIX, No. 58), 9659 + 9660+ 9319, 9662-49665, 9663, 9666, 9667, 9670, 9671, 9673, 9674, 9688 (cf. Pl. XIX, No. 60), 9687 (cf. Pl. XIX, No. 61), 9689, 9692 (ef. Pl. XIX, No. 56), 9695 (cf. Pl. XIX, No. 57), 9696 + 9637 (cf. Pl. XTX, No. 52), 9697 + x + 9927, 9698, 9700 (ef. PI. XIX, No. 55), 9701, 9702, 9903, 9905, 9906, 9907, 9908, 9912 +9658, 9921 + 9313, 9922, 9923, 9925 (ef. Pl. XIX, No. 54), 9926, 9928, 9929. The same, continued. The same, continued. The same, continued. The same, continued. Fragm. of a vase in white calcite stalagmite,27x 10x 2. Nippur III, approximately same place as Pl. 86, No. 86. Inser. 3 col., 1+ 84+2=6li. C. B. M. 9900. Two fragm. of a vase in white calcite, probably stalagmite (glued together), 4.85 x 4.9 x 2. Nippur ILL, approximately same place as Pl.1, No.1. Inser. 4 li. C. B. M. 9648 aand b. Cf. Pl. 37, No. 86, li. 7-5 f. e. Fragm. of a vase in white calcite stalagmite, 5.8 X 7.8 x 1.8. Nippur III, approximately same place as Pl. 86, No. 86. Inser. 5li. C. B. M. 99380. Two fragm. of a vase in white calcite stalagmite (glued together), 4.8 xX 5.5% 1.2. Nippur III, approximately same place as Pl. 36, PLATE. 43 43 43 43 44 44 44 44 44 44 45 Text. 92 93 94 96 97 98 99 100 101 102 103 CHIEFLY FROM NIPPUR. 275 DatTE. En-shagsag(?)-anna. Ur-Shulpauddu. Ur- Enlil. Ur-Mama. Aba- Enlil. [Ur ?]-Enlil. Same Period. Same Period. Same Period. Same Period. Time of Ur-Shulpauddu. Same Period. DESCRIPTION. No. 86. Inser. 3 (orig. 5) li. C. B. M. 9963 4.9998. For the end of the inser. cf. Pl. 43, No. 92. Fragm. of a vase in white calcite stalagmite, 4.5 x9 x 1.6. Nippur III, approximately same place as Pl. 1, No.1. Inser. 3 (orig. 5) li. C. B. M. 9618. For the beginn. of the inser. cf. Pl. 48, No. 91. Two fragm. of a vase in white calcite stalagmite (glued together), 12.5 X 6X1. Nippur III, approximately same place as Pl. 1, No. 1. Inser. 8li. C. B. M. 9616 + 9931 (the former excavated 1890, the latter 1893). Parts of li. 2-7 written also on C. B. M. 9622. Votive tablet in impure bluish gray limestone, round hole in the centre, 2 groups of figures and an inscription incised ; 20.6 x 19.3 X 2.6, d. of the hole 3.2. Nippur X, found out of place in the loose earth along the S. W. side of the Shatt-en-Nil, c. $m. below surface. Between the figures of the upper group 4 li. of inscr., beginning on the right, the last 2 li. separated by a line. Sq. Cf. Pl. XVI, No. 37. Fragm. of a vase in brownish limestone with veins of white calcite, 5.8 X 6.9 X 1. Nippur III, approximately same place as Pl. 1, No.1. Inser. 4 (orig. probably 5) li. C. B. M. 9652. Two fragm. of an alabaster bowl (badly decomposed), 12.2 x 7.2 x 1.1. Nippur III, approximately same place as Pl.1, No.1. Inscr. 10 li. C. B. M. 9621 + 9617. Fragm. of a vase in white calcite stalagmite, 5.1 x 3.3 x 1.4. Nippur III, approximately same place as Pl. 36, No. 86. Inscr. 4 li. C. B. M. 9932. Two fragm. of a vase in white calcite stalagmite (glued together), 8.46.91. Nippur III, approximately same place as Pl. 36, No. 86. Inser.7 li. C. B. M. 9952 + 9699 (the former excavated 1893, the latter 1890). Fragm. of a vase in white calcite stalagmite, 9.7 x 6.3 x 1.6. Nippur III, approximately same place as Pl. 36, No. 86. Inser. 6 li., beginn. of each li. wanting. C. B. M. 9958. Fragm. of a vase in white calcite stalagmite, 3.8 x 5.8 x 1.1. Nippur III, approximately same place as Pl. 1, No.1. Inser. 2 li. C. B. M. 9636. Fragm. of a vase in white calcite stalagmite, 4.2 x 4.5 x 0.5. Nippur 111, approximately same place as Pl. 1, No.1. Inser. 3 li. C. B. M. 9686. Fragm. of a vase in white calcite stalagmite, 8.5 x 9.5 x 2.7. Nippur IIL, approximately same place as Pl. 1, No.1. Inser. 7 li. C. B. M. 9614. Parts of li. 1-4 written also on C. B. M. 9297 (dark brown sandstone), which apparently belongs to the same vase as P]. 45, No. 103 and Pl. 46, No. 110. Two fragm. of a vase in dark brown sandstone (glued together), 7.6 X4.3X 1.3. Nippur III, approximately same place as Pl. 36, No. 86. Inser. 5 li. C. B. M. 9954+ 9924. To the same vase be- longs P1. 46, No. 110. Text supplemented by the follow. two Nos. 276 PLATE. 45 45 45 46 46 46 47 47 47 47 48 48 49 TEXT, 104 105 106 107 108 109 110 111 112 113 114 115 116 117 OLD BABYLONIAN INSCRIPTIONS DatTE Same Period. Same Period. Same Period. A patesi (?) of Shirpurla. A patesi of Kish. A patesi of Kish. Time of Ur-Shulpauddu. Time of Ur-Eniil. Time of Ur-Shulpauddu. A little later. Same Period. Entemena. Entemena. Entemena. DESCRIPTION. Fragm. of a vase in dark brown tufa (decomposed igneous rock), 7.4 xX 7.31. MNippur III, approximately same place as PI. 36, No. 66. Inser. 7 li. C. B. M. 9951. Text supplemented by Pl. 45, Nos. 103, 105 and Pl. 46, No. 110. : Fragm. of a vase in dark brown tufa, 5.4 x 4.9 x 0.8. Nippur ILI, approximately same place as Pl.1, No.1. Inscr.5li. C.B.M. 9623. Text supplemented by Pl. 45, Nos. 103, 104 and Pl. 46, No. 110. Two fragm. of a vase in bluish banded calcite stalagmite (glued together), 4.4 x 6.1 x 0.8. Nippur III, approximately same place as Pl. 1, No.1. Inser. 4 li. C. B. M. 9682 + 9629. Fragm. of a vase in grayish calcite stalagmite, 3.1 x 5.6 x 0.8. Nippur Il, approximately same place as P].1, No.1. Inser. 2 li, {C:.B. M9597 Fragm. of a vase in dark brown sandstone, 13.3 X 7.5 X 1.7. Nippur ILI, approximately same place as Pl.1, No.1. Inser.4li. C. B. M. 9572. To the same vase belongs the follow. No. Two fragm. of the same vase (glued together), 13 x 14.5 x 1.7. Nippur III, approximately same place as previous No. Inser. 4 li. C. B. M. 9571 + 9577. Three fragm. of a vase in dark brown sandstone (glued together), 16.7 X 11X1.5. Nippur III, approximately same place as PI. 1, No.1. Inser. 9 li. C. B. M. 9574+ 9575 +. 9579. To the same vase belongs Pl. 45, No. 103. Text supplemented by Pl. 45, Nos. 104, 105. Two fragm. of a vase in white calcite stalagmite, orig. h. c. 14, d. at the bottom ¢. 16.5. Fragm. 9802: 9.5 x 8.91.9. Fragm. 9600: 8.2 11.8 1.9. Nippur LIL, approximately same place as Pl. 36, No. 86. Inser. (beginn. and end) 3+3=6 li. C. B M. 9302, 9600. Fragm. of a vase in bluish banded calcite stalagmite, inside black- ened, 13.2 x 15.4 x 2.8, orig. d. 17.4. Nippur III, approximately same place as Pl. 36, No. 86. Inscr. § X 4.5, 7li. C. B. M. 9329. Fragm. of a vase in brownish gray calcite stalagmite, 17.1 x 11X1.35, orig. d. at the centre 17.3. Nippur III, approximately same - place as Pl. 86, No. 86. Inser. 10 X 3,13 li. C. B. M. 9330. Fragm. of a vase in white calcite stalagmite, 6.8 X 6.5 X 1.1. Nippur III, approximately same place as P].1, No.1. Inser. 6li. C. B. M. 9655. Two fragm. of a large vase in white calcite stalagmite, outside black- ened, 13.4 14.83. Nippur ILL, approximately same place as Pl.1, No.1. Inser. 2 col., 8+6=14 li. C. B. M. 9463 + 9690 (both excavated 1890). To the same vase belong the follow. two Nos. Fragm. of the same vase, 9.4 X 7.2 X 2.7. Nippur III, approximately same place as Pl. 36, No. 86. Inser. 2 col.,4-+-3=7 li. C. B. M. 9328 (excavated 1898). Two. fragm. of the same vase, 7.1 X 9.9 X 2.6. Nippur III, approxi- mately same place as previous No. Inser. 2col.,5+2=7li. C. B. M. 9219 + 9920 (both excavated 1893). CHIEFLY FROM NIPPUR. QT Puate. Text. Date. DESCRIPTION. 49 118 Dyn. of Kish. Fragm. of a vase in coarse-grained diorite, 12 x 12.2 x 1.6. Nippur III, approximately same place as PI. 36, No. 86. Inscr.6 li. C. B. M. 9918. 49 119 Sargon I. (?) Fragm. of a vase in white calcite stalagmite, 4.8 x 8.4 x 1. Nippur ILI, approximately same place as Pl. 36, No. 86. Inser. 4 (orig. 6) li. C. B. M. 9831. 50 120 Naram-Sin. Fragm. of an inscribed bas-relief in basalt, 52.5 x 39.7 x 8.5. Diar- bekir. Inser. 19.1 X 18.4, 4 col., 2+648+8—24 li. Ca. Orig. M. I. O., Constantinople. Cf. Pl. XXII, No. 64; also Scheil in Recueil XV, pp. 62-64, Maspero, ibid., pp. 65f. and The Dawn of Civilization, pp. 601f., Hilprecht, Recent Research in Bible Lands, pp. 87-89. 51 121 Ur-Gur. Door socket in a black dense trachytic reck, 41 x 25 x18, Nippur III, 12 m. below surface, underneath the W. corner of the S. E. buttress of Z. Inser. 19.7 x 7.5, 10 li. Sq. 52 122 Ur-Gur. Gray soapstone tablet, Obv. flat, Rev. rounded, 12.2 x 7.7 x 1.7. Nippur III, approximately same place as Pl. 36, No. 86. LInser. 5 li. (identical with that on his bricks). C. B. M. 9932. Cf. I R. 1, No. 9. 52 123 Dungi. Dark gray soapstone tablet, Obv. flat, Rev. rounded, 8.3 x 5.6 X 1.6. Nippur X, found out of place in the rubbish at the foot of a’ mound, ¢.1 m. above the surface of the plain. Inser. 6 (Obv.) + 2(Rev.)= 811i. Sq. 53 124 Dungi. Fragm. of a baked clay tablet, reddish brown with black spots, Obv. flat, Rey. rounded, 20.1 x 18.5 4.3. ello. Obvy., 6 col. (23+ 30 + 35 + 22-4 22 4+ 25=) 157 li. Orig. in M. I. O., Constantino- ple (Coll. Rifat Bey, No. 242), copied there 1894. PI. 3 of orig. size. 54 124 Dungi. The same, Rey., 6 col. (214-154 10 4 27 -+ 35 + 18=) 126 li. Copied in Constantinople 1894. Pl. 3 of orig. size. 55 125 Ine-Sin. Two fragm. of a baked clay tablet, light brown (glued together), Obv. flat, Rev. rounded, 12.8 x 6.1 x 2.8. Nippur X. Inser. 19 (Obv.) +22 (Rev.)=41'li. Orig. in M.I. O., Constantinople, copied there 1893. Cf. Hilprecht, Assyriaca, pp. 22f., Scheil, in Recueil : XVII, pp. 37f. 56 126 Bur-Sin II. Baked clay tablet, reddish brown, Obv. flat, Rev. rounded, 20.5 x 19.9 x 3.8. Tello. Obv., 7 col. (parts of col. I-III, VI, VIE wanting, 32+ 19+ 32+31+ 314+30+21=) 196li. Orig. in M. I. O., Constantinople (Coll. Rifat Bey, No. 256), copied there 1894. Pl. 3 of orig. size. 57 126 Bur-Sin II. The same, Rey., 7 col. (part of col. I wanting, 30 + 23 + 21 + 20 + 23 + 15+ 10=) 142 li. Copied in Constantinople 1894. Pl. & of orig. size. 58 127 Gimil (Kat)-Sin. Fragm. of a clay tablet, slightly baked, dark brown, Obv. flat, Rev. rounded,7 X52. Nippur X. Inser. 9 (Obv.) +4 (Rev.) = 13 li. C. B. M. 58 128 Rim-Aku. Fragm. of a baked clay phallus, light brown, h. 14.3, largest circum- ; ference 14.7. Nippur X. Inscr.17 li. Orig. in M. I. O., Con- stantinople, copied there 1893. A. P. 8.— VOL. Xvi. 2 J. 278 PLATE. 59 60 60 60 60 61 61 61 61 61 61 TEXT. 129 130 131 132 133 134 1385 136 137 138 139 OLD BABYLONIAN INSCRIPTIONS DatTE. Ammizadvga. Cassite Dyn. c. 2500 B.C. Burnaburiash. Kurigalzu. [KuJrigalzu. Kurigalzu. [Nazi]-Maruttash. Nazi-Maruttash. DESCRIPTION. Two fragm. of a clay tablet, slightly baked, brown, 11.6 x 10.8 x 3.2. Nippur X. Obv., 3 col. of inser., middle col. Sumerian in Old Babylonian characters, first and third col. Semitic Babylonian in Neo-Babylonian script, Rev. badly damaged, traces of second and third col. The tablet was written c. 600 B.C. Orig. in M. I. O., Constantinople. Fragm. of a slab in white marble with reddish veins, 24.5 x 21 X 6.7. Nippur ILI, approximately same place as PJ]. 36, No. 86. Inser. 2 col., 6+5=11 li. Ca. (C. B. M. 9794). Orig. in M. I. O., Constantinople. Brown hematite weight, ellipsoidal and symmetrical, complete, weight 85.5 grams, length 7.8, d.2.1. Nippur X (June, 1895). Inser. 1.9X 1.8, 3 li. (1. X shiklu 2. din hurdsi 3. dam-kar= “10 shekels, gold standard of merchants ;’’ according to this standard 1 mana= 513 gr.). Sq., sent from the ruins. Seal cylinder in white chalcedony, length 3.4, d. 1.5. Babylonia, place unknown. A bearded standing figure in a long robe, one hand across the breast, the other lifted. A border line at the top. Inser. 9 li. Impression on gutta percha (in possession of the editor). Orig. in the Metropolitan Museum of Art, New York. Cf. Hilprecht, Assyriaca, p. 93, note, Ward, Seal Cylin- ders and other Oriental Seals (Handbook No. 12 of the Metropol. Mus:), No. 391. Fragm. of a lapis lazuli disc, 3.2 x 3. Nippur X,. found in the loose débris on the slope of a mound, and near to its summit (1895). Inscr. 6 (Obv.) + 6 (Rev.) =12 li. Pencil rubbing, sent from the ruins. Fragm. of an agate cameo, 3.95 X 1. Nippur III,same place as PI. 8, No. 15. Inscr. 3 li. Orig. in M. I. O., Constantinople, copied there 1893. Fragm. of an agate cameo, 2.81. Nippur III, same place as Pl. 8, No. 15. Inser. 3 li. Orig. in M. I. O., Constantinople, copied ‘there 1893. Fragm. of an axe in imitation of lapis lazuli, 6.75 x 4.25 x 1.5. Nippur U1, same place as Pl. 8, No. 15. Inser. 7 li. Orig. in M. I. O., Constantinople, copied there 1893. To the same axe belongs the follow. No. Fragm. of the same axe, 4.2 X 3.6 X 1.1. Nippur III, same place as Pl. 8, No. 15. Inser. 4 li. Orig. in M. I. O., Constantinople, copied there 1893. [Kadashman]-Turgu. Lapis lazuli disc, 2.75 x 0.8. Nippur III, same place as Pl. 8, No. 15, Cassite Dyn. Inser. of 51i. (1. [A-naJiuNusku 2. be-la-sht 3. [Ka-dash-man]- Tur-gu 4. a-[na ba]-l [a-ti-sh]t 5. i-[ki]-ish) erased in order to use the material. Orig. in M. I. O., Constantinople, copied there 1893. Agate cameo, hole bored parallel with the li., 2.4 x 1.65 x 0.8. Nip- pur III, same place as Pl. 8, No. 15. Inscer. PingirHn-lil. Orig. in M. I. O., Constantinople, copied there 1893. PLATE. 61 61 61 61 62 62 63 64 64 65 TEXxT. 140 141 143 144 145 146 147 148 149 DatTE. Cassite Dyn. Cassite Dyn. Cassite Dyn. Cassite Dyn. (?) Cassite Dyn. Cassite Dyn. Cassite Dyn. ce. 1400 B.C. Marduk-shabik-zérim. Marduk-ahé-irba. CHIEFLY FROM NIPPUR. 279 DESCRIPTION. Remnant of a lapis lazuli tablet the material of which had been used, 2.12.2. Nippur III, same place as Pl. 8, No. 15. Inscr. 3 li. Orig. in M. I. O., Constantinople, copied there 1893. Lapis lazuli dise, 1.2 x 0.15. Mippur III, same place as Pl. 8, No. 15. Inser. PingirNin-lil. Orig. in M. I. O., Constantinople, copied there 1893. Lapis lazuli disc, 1.2 x 0.15. Nippur IIL, same place as Pl. 8, No. 15. Inscr. PingirHn-lil. Orig. in M. I. O., Constantinople, copied there 1893. : Fragm. of a light black stone tablet, 2.15 x 2.4 0.5. Nippur III, same place as Pl. 8, No. 15. Obv., meaning of characters un- known, Rev., animal rampant. Probably used as a charm. Orig. in M. I. O., Constantinople, copied there 1893. Cf. Loftus, Travels and Researches, p. 236f. Unbaked clay tablet, dark brown, Obv., nearly flat, Rev., rounded, 6.15 X 4.75 x 1.8. MWippur X. Plan of anestate. Orig. in M. I. O., Constantinople, copied there 1893. Cf. Scheil in Recueil XVI, pp. 36f. Fragm. of an unbaked clay tablet, dark brown, Obv. nearly flat, Rev. rounded, 3.86 xX 2.35. Nippur X. Plan of an estate. C. B. M. 5185. Six fragm. of a slightly baked clay tablet, brown (glued together), Oby. flat, Rev. rounded, 16.5 10.5 3. Nippur X. Inser., Oby., 4 col., 39 + 40 + 43 + 15= 187 li., Rev. uninscribed. Orig. in M. I. O., Constantinople, copied there 1894. Baked clay tablet, dark brown, nearly flat on both sides, upper left corner wanting, 5.9 5.21.6. Tell el-Hesy (Palestine), found by F. J. Bliss, at the N. E. quarter of City ILL, on May 14, 1892. Inser. 11 (Obvy.) + 2 (ower edge) +11 (Rey.) +1 (upper edge) +1 (left edge) = 23 li., irregularly written. Orig.in M. I. O., Constantinople, copied there 1893. Cf. Pl. XXIV, Nos. 66, 67; also Bliss, A Mound of Many Cities, pp. 52-60; Sayce, in Bliss’s book, pp. 184-187, Scheil in Recueil XV, pp. 187f., Conder, The Tell Amurna Tablets, pp. 130-134 (worthless 1). Fragm. of a baked clay cylinder, barrel shaped, solid, light brown; h. of fragm. 7.98, orig. d. at the top c. 5.3, at the centre ec. 7.8. Place unknown. Inser. 2 (orig. 4) col., 16 + 22-++1 (margin)= 39 li. Orig. in possession of Dr. Talcott Williams, Philadelphia, Pa. Cf. Pl. XXIV, No. 68; also Jastrow, Jr.,in Z. A. IV, pp. 301-325, VIII, pp. 214-219, Knudtzon, ibid., VI, pp. 163-165, Hil- precht, ibid., VIII, pp. 116-120, and Part I of the present work, p. 44, note 4. Boundary stone in grayish limestone, irregular, 48.5 x 24.5 x 18. Babylonia, place unknown. Figures facing the right. Upper section: Turtle (on the top of the stone) ; scorpion, crescent, disc of the sun, Venus (all in the first row below); 2 animal heads with long necks (cf. V R. 57, sect. 4, fig. 1), bird on a post, object similar to V R. 57, sect. 2, with an animal resting alongside (sim- 280 OLD BABYLONIAN INSCRIPTIONS PLATE. TEXT. Date. DESCRIPTION. ilar to V R. 57, sect. 3, fig. 1), same object without animal (all in the second row below) ; object similar to V R. 57, sect. 6, but without animal (below the 2 animal heads). Lower section: A seated figure, both hands lifted (cf. V R. 57, sect. 5, fig. 1), object similar to V R. 57, sect. 6, last object, but reversed, large snake. Inser. 3 co]., 224 28+11=— 56 li. Sq. Orig. in private posses- sion, Constantinople. Cf. Hilprecht, Assyriaca, p. 83, Scheil in Recueil XVI, pp. 82f. Pl. 2 of orig. size. 66 149 Marduk-ahé-irba. The same, continued. PI. § of orig. size. 67 149 Marduk-ahé-irba. The same, continued. Pl. 2 of orig. size. 68 150 c. 1100 B.C. Upper part of a black boundary stone, 33 x 388 x 20. Nippur. Inscr. 2col.,6+6=121i, Ca. Orig. in the Royal Museums, Berlin. Cf. Pl. XXV, No. 69; also Verzeichniss der (in den Koniglichen Muscen zu Berlin befindlichen) Vorderasiatischen Altertiimer und Gipsabgiisse, p. 66, No. 213. 69 15] Ksarhaddon. Fragm. of a baked brick, yellowish, partly covered with bitumen, 18.5 (fragm.) | Trans. Am. Phil. Soc., N.S. XVIII, 1. JEU BA 10 5 20 L. 7. Erasure of DINcIR, the second character of KA-DINGIR-RA, written by the scribe erroneously before KA. Trans. Am. Phil. Soc., N.S. XVIII, 1. Pl. 25 paul ea ‘al heh ae a f | 3 Sead SSsy3> lewd uy ‘il we _bM ale al Rae 7 u S Lo Ss 64 Trans. Am. Phil. Soc., N.S. XVIII, 1. rf) LY YY DLLs Z LY Lice Wp LY, WY Ui Pl: et \yue JA, Bi Trans. Am, Phil. Soc., N.S. XVIII, 1. Lig ly Lh Li, Ys Lig py Z AIRS FETED ED 82 IE, BY Trans. Am. Phil. Soc., N.S. XVIII, 1. JU, BO 88 Obverse. Lime EEN eee t> WY io | (ree ey WED "VWs Pra Vt SEY? 16 PES IT 4x — a iA mee 7 ==2¢ PALS We ALN YES ETM é ERY > yy i AD Ip } es 20 <> D~ LWY S y My \/ ill KH V = > Kae — V : Trans. Am. Phil. Soc., N.S. XVIII, 1. = = : = Dek oo a ~ ee f aoe : : 10 poagoline, | a HE Ra ee yy Tae By ToT aS 7 ca BS ss Soy ree ey cy ey 4 MH El 20 On the left margin of Reverse are traces of DN y Ses bys On the lower margin of Reverse is tii EUS Pl. 82 ne dedi re eee ES ==ag5 be ph ia eee, ee a Ee At) andasl Thin at ABA It <4 == Bait aes rly dv Ry = PRS ls ese Be. lh 08 Lestat de eae ye te Eye Alt By Dai < d ee = | La EO see ee nee EX, 2) a De ft SN 9% BeBe alti’ ent Sik af ay A, By Ls ewe one ee ppsee Hue a pe HES) BA hia i Noe iet ee rene: i BLY Eps its y, i $30008? ; Meee epee B sf be i fst re oesedd Peg 1 Bid it ype eat tl Bay df Pet ay. a f bsp eh i ey a ee Prep iced es |« Sie | | eee = 7 eee ie eke eS rar Sn sO amieebd| v8 S —~ Pl. 33 panunuo) v8 PH 96 TI 1°90 Pl. 84 Trans. Am. Phil. Soec., N.S. XVII, 1. Col. I. AS V = po Erp VR Bee 5 jay Pb Epa WW VE PY ea BE Tea] +e =z 10 it YS Wr Ya \ (= Ke paz BHHe 10 a8 Wek pe ed Para 20 = j= Z 25 20 Vio eae miu Trans. Am. Phil. Soc., N.S. XVIII, 1. Pl. 85 85 Continued Col. IIL Conte 10 15 20 ye Cit ee ‘ aso 26 ~ = bp D 5 | f 30 NE ENED es : os os PETER i A i Oe par Seg UP nf, Pea, Trans. Am. Phil. Soe., N. S. XVIII, 8. 66 12, BS x Uf, Zl. 1: 9657+ 9607, = YW) 9581, 9901. co eco forms 1. 3 on 3 3 9903, 1. 2: ibid. (9903, 9902). _ _ _ — — - 3: 9657+ 9607-9609, 9581, 9903, (990I, 9902, 9632). - 4: 9609+ 9607, 9581, 9903, 9632, (9902, 9608). - 5: 9609+ 9607, 95814-9643, 9632, (9902, 9608, 9905). - 6: 9609+-9607, 9643, 9608, (9905). - 7: 9609+ 9607, 9643, 9608, (9905, 9634). - 8: 9643, 9608, 9605, (9680, aH 608 9607). ==<- 9605 1.9. ibid., (9633, 9599, @ II 9680, 9703). 9680 11% 1. 10: 9643, 9679, 9605, —f# 9679 (9633, 9599; 9680, 9703). 12x 1. 11: 9591-9679, 9605, (9633, a 9605 9599, 9680). 6 Tk 979 1. 12; ibid. 43 x +A 1. 13: 9591, 9605, ToooT, 350% (9633). EF Ly, 9605 1d x IOOOI Y sl Ub; LUMP - 14: 9591, 10001, (9605, 9633, 9994). . 15: TOOOT, 9591, 9904, (9633) Note:el. 7: The scribe forgot to erase two lines drawn by mistake. L. 14: Eraswre of mu-ac. yy isi ' i Se ee Trans. Am. Phil. Soec,, N. S. XVIII, 8. 86 Continued é + 9904 = 9900 Several: lines wanting eC Be | ae es YU TL id He 1. 16-17: 10001; for 1. 16 cf. also 9900, 9904. 1. rr f. e.. 9635 10 f. e. : 96354-9620. 9 f. e. 9620, (9635)- 8 f. e. : 9620+ 9627 +9635 + 9606. 7 f. e. 9606, 9627, (9604). 6 f. e. : 9606. 9630, 9627, (9604). 5 f. e.: 9604, (9630, 9631,.. 9606, 9917 9639). 4-1 f. e. - 9604, beginn. of 1. 3-1 restor. from 9644, for 1. 4 cf. (9631, 9639, 9634, 9917). 2 f. e.: (9917, 9639). e. : (9607). Pl. 37 Bey iMag Pl. 38 Trans. Am. Phil. Soe., N. S. XVIII, 8. Col. I. x 9646 has 5 perpen- dicular li. a Kk 8614 wy x Only 8615 has this oblique li. EW 8615 9674 has 3, 8614 has 4 angul. li. IO NOTE.—The above text has been restored from the following fragments, COL. L L. 1: frr. 8614, 9646, (9313 9915, 9611, 9923). L. 2: 8614, 8615, 9646, 9921-+-9313, 9115 +9913, 9611, (9674, 9923). L. 3: 8614, 8615, 9913, 9674 on (9313). L. 4: 8614,8615, 9674, 9913, 9662, (9587). L. 5: 8614, 8615, 9674, 9913, (9662, 9587). L.6: 8615, 9610 tenes 9674, 9587). L. 7: 8615, 9610, (9587). L. 8-9: Ibidem. L. 10: (9692, 9642). L. 11: 9696, (9692, 9642, 9689). i ne 9696+ 9637, 9642, 9692, (9689). LL. 13 : 9642, 9637, 9689, 9583, (9642, 9654, 9906). L. 14: 9642, 9654, (9689, 9583, 9906 9637). L. 15 : 9642, 9654, 9318, 9583, 9906, (9689, 9656). L. 162 9642, 9318, 9654, 9906, (9583, 9680, 9656, Pare wL. 17: 9318, 9642, 9654, 9906, (9912+ 9658, 9583, 9659 + 9319). L. 18: 9318, 9642, [written on L. 17] 9906, (99121-9658, 9654,-9659). L. 19: 9318, 9642, (9317, 9651, 9912-+9658, 9702, 9659, 9906). L. 20: 9317, 9318 a (9642, 9702, 9906). L. 21: 9317, 9911+-9651, 9645, (9659). L. 22 9317, 9911, 9645, (9659, 9700). L. 23: 9317, 25, eo, (9628, 9700). L. 24? 9317, 9645, 9628, 9659. L. 25: 9317, 9645, 9628, 9659+.9660. L. 26: 9317, 966049659, (9584, sire, 9300, 9301). L. 27: 9317, 9660, 9584 +9315, 9301, (9300). L. 28: 95844-9315, 9660, 9317, 9301, (9300). L. Ppecctuhoone 9317, 9301, 9660, (9300, 9307). LL. 30? 9584+-9315, 9301, 9317, 9660, 9307, 9300. LL. 31: 9301, 9584-49315, 9660, 930 9300. L. 32%.9301, 9300, (9307, 9315, 9907). es ri ea ee) oo IP, BY Trans. Am. Phil. Soe. N. S. XVIII, 8. 87 Continued 22) *9301, 9907 each 4 horizont. li e@---@ vuswsynrvyse ae LLL SE Lessee °9304 has 3, 8614 has 4 angul. li. 2 9921, re. 9aI5 Sy re. 9667, re. 9662 o I li 9665, 99157 9910 Bare : =

9903, on 9113 the last sign omitted ° 3 x y= 9319 » ae 9319 9611, 9671 E Ww] 31¢ 15 «x 32x F| 9905 35 FAL] 9327: 9319 Oxi ° ~ De teinenn H Ee IPR 9319 ne 53 x and NI on 9905 4 20 FN 9319 ee ean el | Rees 10 Caer pn [= | 9305 34x 7 | 7 x sit 9305, 9654, 9659, 9317 610 8 2 35% 9319 has seven, 7 x . . ; 314 eight perpend. li. isk 9305, © pCR DED TG x ' [SA es za \| 9300 ° ! 4 Ral 21 K---xX 25 4, =-] 23. =” 9651 4 9312 x 4, y, 9656 = EX}es: 9319 Sine ie 3 3 x OPaT 8614 27 x pe es Same varr. as 1. 34 Fe 9319 _ text and margin) ° 9300 has five, a Yo 9665 9319 six angul. li. ae 30 9312 €e) »» = Col. ITI. 40. —. ty CE x UE 1 9307 9922 O ° 9319 29 x : Z EA] €q8 9 Lops Ag1,, 9314) Te. 9650, re. 9625 9315, resp. 9319 Varr. on follow. pl. L. 18: 9610, 9624, 9300, 9305, (9668). L. 19: 9610, 9300 [includes the first three characters ef L. 20], 9305, (9624). L. 20: 9610, 9300, 9305, (9651, 9308, 9685, 9668). L. 21: 9610, 9651, 9300, 9685, (9305, 9668, 9308). L. 22: 9300, 9651, 9610, 9656, (9319, 9305, 9308). L. 23: 9300, 9319, 9656, (9651, 9610). L. 24: 9300, 9319, 9656, 9925). L. 25: 9300, 9319, (9309, 9315, 9925). L. 26: 9300, 9319, 9315, (9309, 9925). L. 27: 9319, 9300, 9315, (9309, 9925). L. 28: 9319, 9315, (9307, 9309, 9300, 9317). L. 29: 9319, 9307, 9315, (9317, 9309). L. 30: 9319, 9307, (9315, 9317, 9309). L. 31: 9659+9319, 9307, (9317, 9315, 9309, 9654). L. 32: 9307, 965949319, 9317, 9654. L. 33: 9307, 9659-+9319, 9654, 9317, (9907, 9314). L.34: 9307, 9659+-9319, 9654, 9907, (9317, 9314). L. 35: 9307, 96591-9319, 9654, 9907, 9314, (9317, 9663). L. 36: 9659+-9319, 9307, 8614, 9654, 9907, 9314, (9663, 9317). L. 37: 9307, 9660++9659+9319, 8614, 9665, 9314, 9312, (9654, 9663). L. 38: 9307, 8614, 966049319, 9665, 9314, 9312, (9914, 9663, 9667). L. 39: 8614, 9665, 9307, 9660+ 9319, 9914, 9314, 9312, (9922, 9667, 9625). L. 40: 8614 [col. III begins], 9665, 9914, 9307, 9625, 9660, 9314, (9922, 9667). L. 41: 9914, 8614, 9660, 9665, 9314, (9625, 9922, 9307). L. 42: 99149320, 8614, 9314-49316, (9660, 9665, 9922). L. 43: 9914+ 9320, 8614, 9314-+-9316, (9646+x+ 9310, 9922, 9673). L. 44: g910+ 9914+ 9320, 8614, 9314+ 9316, (9310 [col. III begins], 9673, 9922). L. 45: 9915-+9910+4 9320, 8614, 9316, (9310). L. 46: 9915+ 99104-9320, 8614, 9316, (9310, 9928). Col. IIL L. F: 9913+ 9320, 9928, 9316, (9G03, 8614). L. 2: 9913+ 9320, 9903, 9916 + 9316, (9928). Trans. Am. Phil. Soc., N. S. XVIII, 8. Pl. 41 87 Continied Col. IIT. , Col. IT. 16 *K 9651 S 9651, Clem 9668 18 x : 9671, P “EHS 9670 9310, resp. 9673 TEN 9310, resp. 8614 I2, Shes 15 x 3 perpend. li. on 9651 FE ne: \ ) ) K rT eI 9309, resp. 9319 e Ve —=|I4 9601, 9319 ji "Varr. on follow. pl. L. 3% 9916-+9316, 9903, (9913, 9928). L. 42 9903, 9913, (9928, 9926, 9916). L. 5: 5903, 9926, (9928, 9913, 9304). L.6: 9903, 9928, (9926, 9913, 9304). L. 72 9903, (9928, 9304, 9926). LL. 8: (9304, 9903, 9928). L. 9: (9304, 9619). L. 10: 9304, (9308, 9619, 9313). L. 11: 9308, (9697, 9619, 9313). LL. 12: 9308, 9697, (9313, 9619). L. 132 9308. L. 14: 9308. L. 15: 9308, 9651, (9668). L. 16: 9308, 9651, (9698). L. 17: 9308, (9668, 9924). L. 18: 9308, (9929, 9927, 9668, 9924). L. 19: 9308, 9929, (9666, 9927, 9924). LL. 20: 9666, 9929, 9308, (9927, 9924). LL. 21: 9666, 9670, (9924, 9927, 9671, 9929). L. 22: 9666, 9670, (9671, 9924). L.23: 9666, 9670, (9671, 9924). L. 24: 9666, 9670, (9671, 9924). L. 25: (9666, 9671, 9670, 9305, 9924). IL. 26: 9305, (9309-++-9924, 9624). L. 27: 9309-+-9924, 9305 [col. II ends], (9624, 9610). L. 28: 9601, 9309+x+ 9924, 9624, (9663, 9319, 9638, 9610). L. 29: 9319, 9309+x-+ 9924, 9601, 9663, (9665, 9624). L. 30: 9601, 9663, 9319, 9309, (5665). L. 31: g60r, 9663. 9319, 9309, (9665, 9312, 9307). L. 32: 9601-+9305, 5663, 9319, (9309+ 9311, 9665, 9312, 9307). L. 33: 9305, 9319, 93C9+-9311, (9665, 9907, 9663). Pl. 42 Trans. Am. Phil. Soc., N. S. XVIII, 8. 8/ Continued Col. FIT. <> 9305 Variants continued. e ~ IS x ° e 36 x | Ee =i 9305 QVAA «05 9319 9311, 9319 —_.... SJ o311 o n 6 ax 9305 == g6o2 Foe 9305 =k 9602 FF 9319 DS o. Ni 9314+ 9316+ 9311 —_ ry e 34 ers 395 | A 9319 ‘ 4316-19311, 9602 “| ‘A 729 AT 70) TUDE TED) Ut 1°99 T 710) *ASAINGQ) vol BRA Lee e ait n pica on ; ait 5 JE, Trans. Am. Phil. Soe., N. S. XVIII, 8. TL 19) LT 1°) ASAI rél be ‘AT 790 TA 29) ep ae a eee en a PP ee or ee F z i Seite ch ae) Bae Pl. 55 Trans. Am. Phil. Soe., N. S. XVIII, 8. > Git BMS Ee DL A 4 £4 Hage ss ie = AES AC AM ARS TA “Ty Bg Lg YING e GLLTO: ZEA GY, U8 GULME ; a 4 > >> aoa ps “1 2 Bd Q Sy Py J =] Laue -<* Ne ROB Yi MF > AT AA > _b “A8JONQ() Ata Trans. Am. Phil. Soc., N. S. XVIII, 8. Pl. 56 126 Obverse. Col. I. Col. FI. Col. IT. Col. LV. Col. V. Col. VI. Col. VIT. aes Fatih zazcll N SD iyo 7 = > ame Y Ly uy, yy Vo SMe iii: | x ee Ny nail Ly Y ul, ) NK (at als xe eres ants] eo) ia = . pe ai p> ss aP = hi wai Away a i ! Ik =) se Dari SS le gon mae 20 soto, Trans. Am. Phil. Soc., N. S. XVIII, 3. JPL, BY 126 Reverse. Col. VII. Col. VI. Col Vz Col. TV. Col. ITT. Col. LT. Col. . : ae BDL gf TIDES ial 5 ‘Shea A AUESusk | Tally MUL TY ae a eas 4 rite Ty 4 Tut (pee ite TSR asi “Col. IV, 11, 12, 6, 19: Col. V, 8, 10, 20: Erasure of the scribe. oat In iceene amy PE 5S Bi dal Wied EAC Ge Reverse. Or 197 BG 4 iy fh AU) | ‘© 1s A A , Obverse. Trans. Am. Phil. Soe., N. S. XVIII, 8. g ay ltl, N.S. XVIII, 8. Trans. Am. Phil. Soc., Y; f DY) yyy yyy, %y ie. df Yj Wh Ue UW MYIDIAI = iy, f Yy i f y y, YY Wis iy eg! Lp Wife jf WY, AS Wy Uap Me “ ; Yee, Uy GY iy) yy My yy Wy ” CG Ly OSMOVA 661 1A Riv MG ~~ Ey ~ ay] a BS A_y Ai (ez “ASL 100 if 7 We Lig Those im JPA, Trans. Am. Phil. Soe., N. S. XVIII, 8. sac NS a SSE SSS » 4, ; Lows 133 i 133 Obverse. : Reverse. 4 = > < pr—~|_J G =- KW er Wet 3 Oxide hala LD Lad ant ES as P

e) WO a Ne BD O i] eZ & fa) A 4 So) Ore i 5 Se > < 7) > 5 O a x = 7© Q fal 1, x x A) I = al Ie Me x 1Q Babylon, Trans. Am. Phil. Soe., N.S. XVIII, 1. Py eeexava 8é PLAN OF THE FIRST YEAR’S EXCAVATIONS AT NIPPUR The Roman numbers indicate the places where excavations were made; the Arabic, the height of the mounds, in metres, above the present level of the canal bed. About five metres must be added to obtain the actual height above the plain. III Kkur—Bint el-Amir (Temple). VII Nimit-Marduk (Wall). De trai pyr eaiet) ee exavali WOMWIIWOE, WASLISIS, JON ILJUIMUES IONE, INCISED, Nippur, Trans. Am. Phil. Soc., N. S. XVIII, 8. 7 ms i ea ¥ ee CNTR Ady * Caio aoa Els nied ae ‘ — Trans. Am. Phil. Soe., N. S. XVIII, 8. PL. XVII MARBLE BLOCK OF LUGALKIGUBNIDUDU, Nippur, in IPL, DOW UL VASE PRAGMENTS OF LUGALKIGUBNIDUDU., Nippur. Trans. Am. Phil. Soe., N. S. XVIII, 3. =ie at ep Sion EE) 1 Trans. Am. Phil. Soc., N.S, XVIII. 8. PL. XIX VASE PRAGMENTS OF LUGALZAGGISI, Nippur, 4 |) IPMLy YORK YASE OF ALUSHARSHID (URU-MU-USH), Nippur, Trans. Am. Phil. Soe., N. S. XVIII, 3. nie an Trans. Am. Phil. Soe., N. S. XVIII, 8. IDIL, SOIT 68 BRICK OF SARGON 1. Nippur oad : St yt i - h my fr Hts Vis te Dey : by i Ht i i “ hibs) ci i i Ta Ihe PL. XXII 6 INSCRIBED BAS-RELIEP OF NARAM-SIN., Diarbexir, SOO@sy IW, Sy BOWIE, Bh a E < oi Ss i E Saar nae Trans. Am. Phil. Soe., N. S. XVIII, 8. IDL, ROGUE 65 BRICK OF UR-NINIB—Nippur, Inscription begins at bottom, Ke er ais! PL. XXIV 68 CLAY TABLET (OBVERSE AND REVERSE),—Tell el-Hesy, Freagm. of a barrel-cylinder of Mardukshabikzerim.—Flace unknown, 66, 67, 68, Trans. Am. Phil. Soe., N. S. XVIII, 8. XXV PL. Trans. Am. Phil. Soe., N. S. XVIII, 8. Inscribed Pebble. Fragm. of a Boundary Stone. ZO, 69, Nippur, IPIG, WOOL BAS-RELIEF IN CLAY WITH AN ARAMAIC INSCRIPTION, Nippur. o) > % O) Z 5} fe) 7) ay E < wo I iy) a isl ae i wi Trans. Am. Phil. Soe., N. S. XVIII, 8. PL. XXVII1 iw DE RNA CODA VASE With KORE FATDE XN) Cr 2000 8) €C—Nippuk: Height, 68.5 em.; diameter at the top, B38 em. Found in an upright position 5.49 m. below the eastern foundation cf Ur-Gur’s Ziggurrat, and 3.05 m. below a pavement which consists entirely of burned bricks of Sargon I and Naram-Sin, It stood 7 m. south-east from ay altar, the top of Which was ¢, 2.40 m, higher than that of the vase, ier Piet Say yea a _ Hany hee Dey Ris) bane ane 1 Gn Trans. Am. Phil. Soe., N. S. XVIII, 8. PL. XXVIII 73 ARCH OF BURNED BRICK LAID IN CLAY MORTAR, C, 4000 B. C.—Nippur. 71 em. high, 51 em. span, 83 em. rise. At the orifice of an open drain passing under the eastern corner of Ur-Gur’s Ziggurrat, c. 7m. below the foundation of the same, and 4. 57 m. below a pavement which consists entirely of burned bricks of Sargon I and Nardm-Sin, View taken from inside the drain, Front of arch opened to let light pass through, ie ty Trans. Am. Phil. Soc., N. S. XVIII, 8. IPL, SOXIDK NORTH-WESTERN FACADE OF THE FIRST STAGE OF UR-GUR'S ZIGGURRAT, Nippur, Trans. Am. Phil. Soc., N. S. XVIII, 3. IPL DOK 73 GENERAL VIEW OF THE EXCAVATIONS AT THE TEMPLE OF BEL.—-SOUTH-EAST SIDE. 1, 6 (8), 7 (g)—Three stages of the Ziggurrat. 1—Kast corner of Ur-Gur’s Ziggurrat. 2—Excavated rooms on the south- east side of the temple and separated from the latter by a street. 3—Causeway built by Ur-Gur, leading to the entrance of the Ziggurrat. 4—Deep trench extending from the great wall of the temple enclosure to the facade of Ur-Gur’s Zigzurrat, 5--Modern building erected by Mr. Haynes in 1894, after an unsuccessful attempt by the Arabs to take his life. ya Teter? chy i CAE Pian Hk ah} XVIII. PartIlIl. Plate XIV. Transactions Amer. Philos. Soc. Vol. ——<—<— Representative types. Rana tea Faittsosobbonobis wan Rees Z Se = SS Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XV. , 1 yee : ae! i deacons \, “| a Whe eal Iai OD \\Neroarn | ame); e th - Wh pall \a . as SS 24 Tortricina. Peropoda. Calamariine. Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XVI. NY \ Ni oes ers \ SSS See e, ee oO Oe ee ———— CS SSS = a Ee OS Ga ec ee De ox Ae pul i232 2 - me Catal =— Hy) J ) fi ; ‘) ) Mi ne ge 4) Hf Mi Mh i, Mi iN) ! MW 4 N | ihe i Mh i) i) | Ht i ; NK) yh a s = SSS EES = = Colubrinz. 1 CCM Co iy yee == SEB — 5 . N Th) =) he) ea ee SO XVII. Part Ill, Plate XVII. See Narr ae fe : =a SS i se aA 5 ; iat —— a sas nutuue tates a peta SOS Etc ee HA E 3 se a iid ae oe I 5 : Gai tis i Hata hes : _ picatterentet ter nads SUI Ss ees E BEE SE SETS ES OES s é S SSS ——<—, S = s) : 73233 : : S Spt hts : iS ge , = =e ————— £ SSS aa = eS So ee re SSS Seana an NS eins — ¢,! fl ae " Se ae i ya : ga SE qe HS gh 2a". OT TT pe ee re pp oes Tranasctions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XVIII. Colubrine. Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XIX, 8 & i~) 2 SS i) eg (Se) aS = Z ait = 34 SV —— Ss ——____—J es 5 & Loe = IE bye Whe eS OO ioe a = 7a a ails PSD DE = =sS AE, te SM He Ws . SS Rete SSE, << Sst St <2 eS NS ain = ie : a $s oe esa on, OO ———————— eas on: On gt ey aia ; pee ea = Transactions Amer. Philos. Soc. Vol. XVIII. Part III]. Plate XX. “a y Kt Aa ie Kies eae Hi ae ae: NA! Ae Hy Mil CaN ih a Mi yi PANG \ : Ne W Wa FA i Kv el rae ——_ a5 — 3 7 Sista AO | PORTAL AN ANGE YY Wo | rial JSON EANANY) | SAA HANNAN RN Hh Ca HUN) HI Lal | i iy | l/ WAG WY Colubrine. or a c. Vol. XVIII. Part III. Plate XXI. Transactions Amer. Philos. So CS) SS eS Colubrine. Natricine. SS SS == —— —== eee Mi | HANH | i Hh Nt i Hh = SS —— SSS SS HN AVN ———<———= ———— aS— SA ——— i SSS 3555 —s : SS LLL, SSS —S— —S= > Svea Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XXII. EZ = Yo a= =F A —— = SS SAMI ND SSS 2 = 223225335 = Ast Sess Ss => == =—S —SS— —— ——s =——s Natricine. Sm Nerd erage a => Se SSS HWA tf Wn i vlh) AK HIS RAR i) \ HAAR ti si [Se Ree L— Transactions Amer. Philos. Soc. Vol. XVIII. Part Ill. Plate XXIII. = «© <—_. a 2 are i DD — Se a ——— Rae = <= SS — == fi ( ie " HANH Ni Mi HW DER WA == vil i ANN ANAM WAIN TAANY I i My i ih Natricine. Homalopsine, is aes) re Transactions Amer. Philos. Soc. Vol. XVIII. Part Ill. Plate XXIV. ES GE Cg =< Erg GS ai x l ( a = = a ( Co = = fi x [i 4 = — a —— —————__§ ————— < = = — SS Xenodontine. Dromicine. = : is a ey 3 By ela a4 ql 2 Zz = Se —= EY i =e << — ——— ———— — SS = ———= —— Transactions Amer. Philos. Soc. Vol. XVIII. Part Ill. Plate XXVII. Re aS } FE ae ALAN are sea co un Be cea tit ( | igi ul Mu i\ | NAO Xenodontine. Dromicine. Scytaline. Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XXVIII. i, M uy ee = Sa Orr iathe pr a= gas 2 = SS ESS SS SS SS Erythrolamprine. Scytaline. Transactions Amer. Philos. Soc. Vol. XVIII. Part Ill. Plate XXIX. Dromicine. Leptognathine. Dipsadine. Erythrolamprine. Transactions Amer, Philos. Soc. Vol. XVIII. Part III]. Plate XXX. hee wn Minot iden Mii Hi iN Wits —— ————S—S SSS s SS E SSS te —s = Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XXXI. —Z> EEE —= Sr —— ——S== —— SSS ——= = —= = 5 — SSS ————= ——s —SS 10 —— ———S> —— ——=——= S—S= — SS ——_!>= = ——— =—> = = === = —S —S ——_ SS SS —= == === ST —— —— SSS —— SS SSS — ——. —S SS =F: SSS = S= ———_—— a Proteroglypha. Platycerca. Transactions Amer. Philos. Soc. Vol. XVIII. Part III. Plate XXXII. Sa — Ss Ss aS ———— == Soe == — = =e are? Hs = SEES 3 ‘ : = tS roy 0 aeSeonenls Hu. ee aes LESS 5: = r xue > } ) 333 HEN, med ‘ a 3) ss