CORRESPONDENCE.

1. THE BIBLIOGRAPHY OF AFRICA.

December 5th, 1887.

SIR,-In the last issue of the Journal, I announced the forthcoming publication, by the Roman Catholic Missionaries of Senegambia, of a Dictionary of the Susu language. I was then unaware of the fact that this book had already been published two years ago, and I have only just found it out from a German Catalogue of second-hand books. The Dictionnaire français-soso et soso-français, to which are prefixed a grammatical sketch and a collection of common phrases, will prove a very valuable Handbook of this language, which is spoken along the coast between the Rio-Nuñez and Sierra-Leone. The author is the Rev. P. RAIMBAULT, and the work, though printed in Paris, has been issued by the Mission du Rio-Pongo, Vicariat apostolique de Sierra-Leone, 1885.

What the said Missionaries were going to publish was really a practical Grammar of the Bambara language, which has now been issued (Eléments de la Grammaire Bambara, etc. 1 vol. 16mo. vii. and 218 pp., Saint-Joseph de Ngasobil, 1887). It contains numerous exercises with lists of words, and it is followed by some texts with a Bambara-French Dictionary: this is the most complete and elaborate work ever published on that interesting language.

I must also quote here a little work, issued by the same Missionaries in 1880, which is not noticed in Cust's Modern Languages of Africa, and which would prove very useful to Englishmen, because it contains an English translation of all words and sentences; its title is (in French and in English) as follows: Guide of the Con

versation in four languages, ENGLISH-WOLOF-FRENCH-SARAR, 1 vol. 32mo. 329 pp., Saint-Joseph de Ngasobil, 1880.

The Secretary of the Royal Asiatic Society.

CAPT. T. G. DE GUIRAUDON.

2. NOTES ON AFRICAN PHILOLOGY.

December 20th, 1887.

SIR,-Amongst the Notes contributed by the Hon. Sec. to the last issue of the Journal, I read as follows:

The

"Vocabularies of the Hadendoa and Beni Amir. Hadendoa is a Dialect of the Bishári language, of the Hamitic group (see page 126 of Cust's Modern Languages of Africa, 1883). The Bani Amír are wrongly entered as a Dialect of the same language, but the Vocabulary shows that the language is Semitic, and akin to the Tigré of Abyssinia."

So, if I understand rightly, we are told by Dr. Cust himself that he has been wrong in entering the Beni Amír as a Dialect of the Bishári language: we shall see hereafter that this statement should really be understood in a way quite different from that suggested by the phrase quoted above.

I must observe, in the first place, that Bani Amir, or, more correctly, Beni Amer, is a plural ethnic tribal name (Hebrew Amrim), meaning "Sons of Amer," and I fail to understand how the "Sons of Amer" could be styled a Dialect. We could not say that the Dutchmen are a Dialect. With regard to these Beni Amer, as the Vocabulary alluded to has not yet been published, I must postpone my opinion on the question whether the language is Hamitic or Semitic. But both suppositions are possible for some of the Beni Amer, who are of Tigréan descent, have preserved their original Semitic dialect, while the rest of them now speak a Hamitic dialect (see W. Munzinger's Ostafrikanische Studien and Vocabulaire de la langue Tigré). Therefore, if Dr. Cust confesses himself wrong in entering the Beni Amer as a dialect of the Bishari language, he would have rightly corrected himself by entering their name as that of a tribe speaking partly a dialect of

the Tigré language (Semitic) and partly a dialect of the Bedawye language (Hamitic). In other words, his entry is right, though incomplete, as to the name of the tribe, but it is quite wrong as to the names of dialects and languages.

I come now to the so-called Hadendoa dialect of the so-called Bishári language.

The language, which these people who speak it call to' Bedawye, i.e. the Bedawye (see Munzinger, Reinisch and Almqvist), and to which we have, therefore, no ground at all for applying any other name, is spoken, according to the best authorities, by the Hadendoas, the Bisháris, the Halengas, the Amarars, the Ababdehs and a fraction of the Beni Amer. To call this language by the name of any of these tribes, is exactly as if we were to call the French language the Auvergnat, and we should only aggravate such a mistake by speaking further of the Britton or Picard dialects of the Auvergnat language. We can only speak of the dialect of the Bedawye language, as spoken by the Hadendoas or the Bisharis.

Both Hadendoa and Bishari are but the names of tribes speaking, together with the others mentioned above, one and the same language, and none of these appellatives can be applied to the common language, the right name of which we know perfectly well, as already stated. That all these tribes speak a common language with some dialectal differences (which, after all, are mostly mere differences of pronunciation), this fact is beyond any doubt. But that is the only difference we can trace. We cannot speak of dialects in the true sense of the word among uncultured tribes. Very often the language becomes modified from place to place. It is very difficult, not to say quite impossible, to state where a socalled dialect begins and where it ends, and we can only say where a language, in one or other of its dialectal forms, begins and where it ends. Therefore, when an author tries to separate such dialects one from the other, he runs the risk of becoming quite unintelligible and of heaping mistakes on mistakes. A few more quotations will more fully illustrate what I mean to say.

In his above mentioned work (p. 159-160), after having stated, though without any ground, that there are five-I could as well say fifteen or seventy-dialects of the Fulah language, Dr. Cust goes on quoting: "Faidherbe admits that his Grammar is of the dialect of the Toucouleur, or Futa Toro, . . . It presents several differences from pure Fulah, . . . " and further: "Baikie observes that the language was spoken in its purest form in Futa Toro . . . ." VOL. XX.-[NEW SERIES.]

10

It seems to me that all this is so illogical and self-contradictory, that though it reads like statements of facts, it really conveys no meaning at all.

How can one speak of the purest form of a language which has no literary standard, the only available one: I mean no true indigenous literary standard, as I cannot consider the translations of the Bible made by some missionaries otherwise than as an artificial literary standard. But, if this language is spoken in its purest form in Futa-Toro, how can this purest form present several differences from pure Fulah? And in what part of FutaToro is this purest form to be found out? During more than three years I spoke myself exclusively the Pul language at many different places of Senegalian-Futa (Futa-Toro, CentralFuta and Futa-Damga), and every where I found some dialectal changes but I have no term of comparison to say whether the purest form was spoken at Gourik (Futa-Damga) or at Podor (Futa-Toro); I can only say that the dialectal forms spoken by the Bosseyabes and other tribes of Central Futa are perhaps less mixed with foreign words than the others. In fact, there are two great dialectal forms of the Pul language, which are spoken in two separate countries, Senegalian-Futa and Futa-Dyallo: elsewhere, the Fulbe being more or less scattered amongst foreign populations, their language has become mixed and altered in various ways, and it is quite impossible to speak of any dialectal classification.

:

Returning eastwards overland, I come to what Dr. Cust calls "Nile sub-group," and here I find in his Bibliography :

For the so-called Moru dialect of the Bari language, as illustrated by Col. E. Long, is not at all a dialect of the Bari language, with which it has not even two words in common. On the other hand, the Moru dialect looks so very much the same as the Nyangbára language, as illustrated by Morlang, that it may be asserted with

all certainty that both Nyangbára and Moru are but dialectal forms of one and the same language.

I would not myself venture to give any new complete classification of African dialects and languages, as I consider it to be impossible for the present, and, in making the few preceding remarks, I had only in view to point out the difficulty of the subject in the present state of our knowledge.

CAPT. T. G. De Guiraudon.

The Secretary of the Royal Asiatic Society.

Note to the above by the Hon. Secretary.-All contributions to our knowledge of these imperfectly studied African languages, made by specialists, who, like our correspondent, have actual personal acquaintance with the subject, are of extreme value, and we thank Capt. de Guiraudon for his interesting communications, and we hope to hear from him again.

3. THE MIGRATION OF BUDDHIST STORIES.

MONSIEUR, Dans son important article sur la Simhâsanadvâtrimsikâ (Ind. Stud. xv.), Mr. Weber ne croyait pouvoir rattacher de près les fragments d'une recension Mongole, connue sous le nom "Histoire d'Ardshi-Bordshi Khân," aux textes des recensions Samskṛtes. Une traduction Persane, faite pour la première fois du temps d'Akbar sur un texte Indien et remaniée plusieurs fois aprés, nous fournit des données précieuses pour le rapprochement des textes en question. Il existe de cette version Persane une traduction française du baron Lescallier (Le trône enchanté, New York, 1817, 2 vols. 8vo.), aussi infidéle, que rare (ni Benfey, ni Weber n'ont vus cette traduction). L'Introduction nous donne et l'histoire du père de Vikramâditya-Gandharva-sena, transformé en âne par une malédiction d'Indra, et l'histoire du cadavre flottant. Le récit de la 7me statue présente certaines analogies avec l'histoire du chasseur et des perroquets, pour laquelle nous trouvons une parallèle très rapprochée dans la littérature orale Indienne. Le récit de la 10me statue nous donne une version de l'histoire de Naran Dâ Kinî.

Cette petite notice a pour but de signaler l'étroite affinité entre la récension Buddhiste Mongole et une des recénsions Indiennes. Je compte, sous peu, donner une analyse détaillée de la version Persane d'aprés plusieurs MSS. de Londres et de Paris.

SERGE D'OLDenburg.