In the Mongolian language we often find two, three, or even four verbs in succession without a single conjunction between them. When analysing a continuous chain of verbs, it is best to consider two consecutive verbs at a time, for the circumstances in which such a combination is made can be brought under the following three headings:—

1. Sequential—two verbs denote actions in sequence, as, for example, nadur irejü ögüleṙün, He came to me and said …;

2. Adverbial—the first of the two verbs is merely adverbial, usually expressing the manner 1 in which the action denoted by the second verb is performed, as, for example, uyilan jor010D;cibai, They went away crying;

3. Auxiliary—the second of the two verbs is auxiliary to the principal verb, as, for example, bi čimayi alaqu bui, I shall (lit. am) kill thee.

Of these three cases the most interesting is undoubtedly the last, i.e. the auxiliary use of a limited number of verbs with certain welldefined meanings, an exact parallel being found in other Eastern languages.

The vicissitudes suffered by this famous architectural treatise, and especially the sources of the splendid re-edition published in 1925 were the subject of an earlier article in the Bulletin (Vol. IV, Pt. Ill (1927), pp. 473–92). A happy chance throws new light on the 1925 edition and allows me to add this note to what was said before. It is the presence in England of a sixteenth century manuscript copy of a part of the Ying tsao fa shih. If the editors of the 1925 edition had had access to this, they would doubtless have turned to it for data in their efforts to reconstruct the lost Sung original. At all events, the drawings it contains are of great interest to students of Chinese decorative design, for reasons to be mentioned presently.

The Lau language is spoken by the inhabitants of the “ artificial islands ” in the lagoon off the north-east coast of Big Mala, Solomon Islands, and by the people who live on the natural island Ngwalulu (Manoba, of the Admiralty charts) which closes the northern entrance to the lagoon. The southern end of the lagoon is known as Ataa Cove, and the language of the islanders in that part of the lagoon has a closer affinity to the languages of the hill peoples of the mainland than has Lau proper. In Port Adam, Little Mala, there are two villages whose people also talk Lau. they being immigrants from Ataa Cove. The hill languages of the north end of Big Mala use certain sounds which do not appear in Lau. These sounds are ngg, i.e. ng in English “ finger ”, nd instead of pure d, mb instead of pure b. Also the Lau h appears as s or th.

With a magnanimity which is singularly rare these days Mr. H. Kevorkian, a curio-seller of New York, lent me a year ago his exceedingly precious and valuable MS. of an anthology of Persian poetry, giving me full permission to utlize or publish it, and I take this opportunity of thanking Mr. Kevorkian for his liberality. The MS. is entitled Mu'nisu'l-Ahrr fi Daq'iqi'l-Ash'r and contains copious extracts covering all branches of Persian poetry, from the elegy, the ode, and the double-rhyme poem to the quatrain and the unit, quoted from two hundred Persian poets whose names are found on the first folios of the MS. and are the reproduced lower down.

The Material for the study here presented was collected during a short stay at Marau Sound in 1927, in pursuance of my research work. Bishop Patteson, of the Melanesian Mission, published grammatical notes of the language of Marau Sound at Auckland, New Zealand, about 1860, together with a short catechism, a translation of the Apostles´ Creed and the Lord's Prayer, and a list of words. This material was edited by von der Gabelentz, and was published in his Melanesischen Sprachen, Leipzig, 1873. I have worked through the material and corrected it where necessary. The two pieces of translation, Creed and Lord's Prayer, which appear in von der Gabelentz. are too faulty to be inserted here as specimens of the language, and I had not the opporetunity of correcting them.

Bishop Patteson had several men from Marau Sound with him at Auckland, among whom, according to the information given to me, were Tootoo, Waaro mae, Waihunu, Nini pua, Porike, Porasi (who died at Auckland), and also Vouvete of Kaoka, a village on the mainland near by. The large percentage of San Cristoval wordsand grammatical constructions which appear in Bishop Patteson's material, as quoted by von der Gabelentz, is probably due to the influence of Taroaniara, a San Cristoval man, who acted as guide on the visits to Marau Sound, and whose wife was a Marau Sound woman. Taroaniara was with the party at Auckland, and evidently acted as interpreter. He may even have been commissioned by the Bishop to render certain stock pieces, e.g. the Creed, into the language of Marau Sound. This supposition would account for the presence in the translations of words like tahi, to live, tataro (San Cristoval dadaro) for “ cross ”, and for such grammatical constructions as ia, personal article in San Cristoval, instead of a, the Marau Sound usage. Also it would account for the use of ni, demonstrative article, and ia, personal pronoun 3rd singular. However, there is good internal evidence that the work in the main was done by the Bishop.

The results of cultural anthropological study among the people of the Sudan and Sahara do not always yield definite and conclusive results with regard to ethnic groupings and cultural contacts in the past. On the other hand the Sudanese migrations and contacts of the past as portrayed in the bizarre, and often allegorical stories told by the Sudanese people themselves with regard to their origins and antecedents, often seem so much open to doubt in European eyes, that an enquirer is apt to think that the ethnographical record of most of the northern half of Africa is so blurred with contradictions, and so confused by inaccuracy, as to be undecipherable.

The two famous monuments known as the Orkhon Inscriptions, erected in honour of the two Turkish princes, Kül-Tegin and his brother Bilgä Kagan, though mentioned in Chinese histories, remained forgotten and ignored down to quite recent times. They stand near the Lake Kocho Tsaīdam, to the west of the River Orkhon, about 50 miles north of the monastery of Erdentso (the site of the ancient city of Kara Korum), and about 25 miles to the north-west of the ruins of the capital of the Uighurs, Kara-Balgassun. They are two great square monoliths, which originally reposed in two sockets. Both these sockets are still in their original position, but the monuments have been thrown down, with the result that Monument II is in four pieces. They contain long Turkish inscriptions in Runic characters on three sides, and on one side (the western) the inscription is in Chinese. The Chinese inscriptions are very beautifully cut; the Runic inscriptions, which are also probably the work of the Chinese stonecutters, are less elegant. It may be mentioned that the Chinese inscriptions differ entirely in their contents from the Turkish.

Of the five Temazheg poems which follow I have only attempted the translation of four. The fifth, though the vocabulary is comparatively simple, is so elliptical in construction that a translation is impossible. The same applies in a lesser degree to the others, the translation of which is therefore somewhat free.

Tuareg poets, besides being fond of topical references which are incomprehensible to strangers, are inclined rather to flatter the intelligence of their hearers than to err by being too explicit. Grammar and syntax are freely sacrificed to metre and effect.

WÖrter vom Stamme kommen mehreremal im Ḳurān vor. So 20, 56: Speiset und weidet euer Vieh. 28, 23: die Hirten. Die Weide heisst: 79, 31.87, 4. Im Sinne von “ hüten, beobachten” 23, 8.70, 32: und die ihre Verpflichtung einhalten, beobachten. 57, 27: aber (das Wohlgefallen Gottes) beachten sie (die Christen) nicht der richtigen Beachtung gemäss. In 2, 98 wird denjenigen, verboten zu sprechen, dafür sollen sie schaue auf uns, beachte uns, behüte uns gebrauchen. In 4, 48 wiederum wird denen der Vorwurf gemacht, dass sie zu sprechen pflegten; besser wäre für sie (v. 49), wenn sie statt dessen sagen würden. Da die beiden verschiedenen Wörter verwandte Bedeutung haben, in fast synonymem Sinne gebraucht werden können, so ist nicht einzusehen, weshalb Muḥammad den einen Ausdruck zu verwenden befahl, den anderen hingegen, der ihm offenbar missfiel, zu verwenden verbot.

The following poem was copied for me in 1912 by Muhammad bin Abubakar bin ‘Umar (Muhamadi Kijuma) of Lamu. The author was one ‘Umar bin Amin, a blind Sharif of Siu, said (in 1913) to have died “ about fifty or sixty years ago ”. This information was furnished by Ahmad bin Abubakar bin ‘Umar es-Siwi, of Mambrui, who added that ‘Umar recovered his sight on completing the poem ! His name, by a not uncommon practice, is spelt out in the final stanza. So far as the poem shows any linguistic peculiarities, these belong to the Lamu dialect, which differs considerably from the poet's native speech of Siu. This may be due to its passing through the hands of Lamu copyists, or, more probably to the deliberate adoption of what was considered to be a more literary form of language.

Whether or not the text has been corrupted here and there, I find some obscurities which I have been unable to translate. These I have left with a query, in the hope that some reader in East Africa may be able to elucidate them. Some allowance, moreover, has to be made for poetical licence, in view of the difficulties of this very artificial kind of composition.

The following stories are the first two of a Mabale series. Mabale is a Bantu dialect spoken by the Ba–mabale (sing. Wa–mabale; both abbreviations of moto wa Mabale, a Mabale-man and batu ba Mabale, the Mabale people).

The Bamabale live on the right bank of the Congo River, up but especially down Nouvelle–Anvers.1 They also occupy the immediate hinterland and are even found in a few places on the left bank, opposite the same central town.

The Bamabale belong to the group of tribes we usually call Bangala. Amongst them are the Iboko, who also live at Nouvelle–Anvers, and whose ancestor Mata-Boiki (i.e. grandson of Boikr) has become famous, since his encounter with Stanley; further, the Mbenga, Motembo, Losengo and Boloki (between N.A. and the mouth of the Mongala); the Ndobo and Mbundzi (down N.A.), all on the right bank. The BaFәtә (near Lisala) and the εlεku (at Lolanga and near Coquilhatville); the Libinza, Ewaku, and Likoka (in the hinterland, on the Ngiri).

bakεndεkε (1) dzisolo. (2) bakomi (3) omolako, They went fishing. They arrived at a fishing-camp,

balambi (3) bilεtibakalε. (4) ηkumbajoti: they cooked food so they were going to eat. The tortoise he so:

“ tobomba (5) bilε; tokεndε (6) boodzisolo. “ let us put aside the food; let us go first to the dzisolo.

wautaka (7) biunansalatokalε.” (4) (when) shall have come out we with hunger we shall eat.”

lokulukokoaambi(3); bakwedzidza(8) ηkandu (9) obwato. (10) The pheasant agreed; they put the hurdles into the canoe.

mwanawalokulukokoaikadzi (11) omolako. The little one of the pheasant stayed in the fishing-camp.

badziηgi (3) dzisolo, baabi (3) nsolo, i'idzi They surrounded the dzisolo, they cut the grass, they were

njε (12). finished entirely.

In the Linguistic Survey of India the language spoken in the Western Panjāb is called Lahndā. Previously it had, in India, not been recognized as any independent form of speech, the many local dialects there spoken—Mūltānī, Sirāikī, Hindkī, Jaṭkī, and so on—being looked upon merely as so many dialects of Panjābī. Panjābīs themselves had no general name for this group as a separate entity. When they wished to express that idea they employed a periphrasis, such as Lahndē-dī bōlī, or “ the dialect of the West ”.

European scholars, however, had by that time long recognized the fact that a general name for the whole group was needed, and more than forty years ago one of the first describers of the language, Mr. Tisdall, named it “ the Lahindā ”, i.e. Lahndā, “ dialect.” I am not especially enamoured, myself, of this name, but as it had not been challenged for some thirty years, as it was not inconsistent with English idiom, and as no better name had been suggested, I employed it in the Survey.

The Japanese language contains a number of words that are pronounced differently when standing in the attributive position. The following is a list of such words, taken mainly from the literary monuments of the eighth century A.D.1

(1) e:a. ame heaven, sky: ama-hire (sky—shawl) cloud. ame rain: ama-giri2 (rain—mist) rainy mist. fune ship: funa-de ( < *funa-ide ship—departure) departure of a ship. ine rice-plant: ina-muširo (rice plant—mat) a pillow-word used for qualifying šiki (dense, heavy, frequent) and kaha (river) which are homonymous with šiki (to spread) and kaha (skin).

Persia is the land of imamzadehs or “ the tombs of saints ”. Nearly every little town in this country is the fortunate possessor of one or more imamzadehs, to which hundreds of pilgrims come all the year round from far and near. The tomb, or very often a simulacre of the tomb, as in the case of a good many imamzadehs no saint lies buried in it, forms its chief part, its sanctuary. There is hardly a sign of its orientation, except that the saint is buried with his face in the direction of the Kibleh. The shrine in which the tomb is placed is generally decorated by a dome. To this the imamzadehs of theprovince of Khuzistan in the south-west of Persia form a remarkable exception. They have all a tall cone, surmounting the flat roof of the shrine. This cone resembles a pine-cone with its series of shallow niches, which forms its only decoration. It ends in a short pointed metallic rod adorned with four metallic balls. This gilt ornament is common to most of the mosques and Muhammadan shrines.