the Canaanite or Palæogeorgian names in Sk are transliterated in Hebrew by Sh. Thus there is a confirmation of the common origin of the mythology of early civilisation, which is discussed by Mr. Campbell. Upon this topic we shall have to seek for origins among the Agaw nations rather than among the Aryans. What can be learned safely from Herodotus is, that there were in his day dark populations in the Caucasus, that they practised circumcision, and that a language like Egyptian was spoken. We may consequently admit that there was then a resemblance between the populations of the Caucasus and those of the Nile region.

Looking further, we shall find that neither the Caucasian nor the Nilotic region is to be regarded as a sole centre of the populations speaking Caucasian or Nilotic languages.

True it is, the Egyptians (Southern or Ude) cannot be traced further; but the earlier members in each group, even in the present state of the investigation, I can identify elsewhere.

The Agaw or Abkhass is defined in West Africa, in the Gadaba of India, and in the Rodiya of Ceylon. It is in the New World, in South America, that we have its greatest present extension, covering Brazil and Guiana as the Guarani, Omagua, the Movima, and the Sapiboconi.

Without extending the chain of evidence, these Agaw groups are sufficient to show that the Caucasian and Nilotic regions are not sole or chief centres, but only local centres or ganglia of large migrations.

If this is the case with the Agaw it must be so with the Egyptians, which is far later. The Agaw migration or conquest must have preceded the Egyptians, but there was an earlier member in all the nuclear regions, still identifiable, ethnologically or linguistically, in the Nilotic region. This is represented by the family of the Gonga lauguages, Kaffa, Woratta, Yangaro, and Dalla, whose affinities are with the remarkable Mincopie languages of the short or dwarf dark races of the Andamans, mythologically or linguistically recognisable in so many centres of refuge in the Old and in the New World.

Unless, therefore, we assume that all races had an African origin, including the Mincopie and the Agaw, and their intervening or accompanying members, we cannot attribute an African origin with any certainty to the Egyptians. The fact of any portion of their populations being dark is not in ethnological consideration a point in favour of an African or tropical tendency, because, in the prehistoric as in the present epoch, dark races can be found in temperate or cold, as well as in tropical districts.

It must consequently be regarded, at least, as an open

question, whether the Egyptians moved from south to north, or from north to south. In the supposition that there were two Egypts or Mitzraim, one in the Caucasus and one in Africa, we get a possible solution of some prehistoric or protohistoric problems.

We must first accept a harmony of ethnological conditions, that in Caucasia and in Africa there were not only dark races, but at an early period Mincopies, represented by pygmies in legend, in both regions. We have also two Agaw lands, possibly two Havilah.

In examining the peculiarities of Egyptian grammar, in Coptic, many of these are found in Agaw or African; in Ude they are found in Abkhass or Caucasian. In the latter case, they are considered as examples of Caucasian grammar. In Ude are many Abkhass words. Lest it may be imagined that in these instances the Egyptian influenced the Agaw, instead of the Agaw influencing the Egyptian, we may hereby seek our test in the American Agaw, the Omagua, or Guarani.

In Egyptian, we have numerous monosyllabic roots, and of these many are apparently of the same form, but differenced in pronunciation, so as to distinguish the various significations. This we find repeated in Ude and Abkhass, but what is more to be regarded, we find it in the South American Guarani.

In

There is, so far as appears, no such thing as a Caucasian grammar, and no such thing as an American grammar. the case before us there is an Agaw grammar. The Caucasian peculiarities are not local, and in South America they cannot be regarded as Caucasian, while Guarani or Omagua is just as much Caucasian grammar as American grammar.

The fact is, ethnological evidence will force on philologists a new system of classification, which can no longer be by localities but by race. When a particular race, as the Agaw, the Semitic, or the Aryan, has influenced the grammars of other races, the effect may in any district appear to be local, but it must assuredly be due to race. Philology is much more dependent on physical researches than has been supposed, as physical researches have nearer relations to philology than there has been any disposition hitherto to admit.

On finding an influence of Agaw grammar on Egyptian, we must be prepared to allow that this will not be the only mental influence, and not the sole propagation or development of aggregate and continuous thought, and we shall have to seek in the mythology and folk-lore of the preceding races much that has been hitherto regarded as exclusively and generically Egyptian and African.

If there is at first a confusion in accepting a north to south

migration of the Egyptians, instead of that from south upward, it may nevertheless not appear so unaccountable, if we regard the annals of later migrations, those of the Hebrews. Here we have, first, an alleged migration from north to Egypt, and then from Egypt in the south by the north to Canaan. Here is a race white in its main elements, but showing a decided tendency, in some cases, to the hair and features of the North Africans, and this race speaks a Semitic language, which has affected the whole of North Africa.

Havilah is in the oldest collections a double name. It is that of the son of Cush, the brother of Mitzraim, in the Book of Generations; but it is the land in which the river Pison or Phasis flows, in the description of Paradise. Havilah I believe to be Khavilah, the land of the Agaws. If, according to a prevalent opinion, we accept it as Colchis, then it is undoubtedly the country of the northern Agavs, Akhaivi, or Avkhass. The northern Havilah would be the Agaw land on the Nile.

When we have to deal with Paradise, after accepting Pison in Havilah, Hiddekel in Assyria and Euphrates, there has ever been a stumbling block in the river that compasseth the whole land of Ethiopia. With a Cush, Mitzraim, and Havilah in the Caucasus, the rivers can be better accounted for, and in conformity with the Chaldæan or Caucasian type of the legends. Many of these cannot, as commonly supposed, have originated after the captivity in Babylon, but they are in this valuable record in a most ancient form.

In this consideration, we find at once an explanation of the remarkable resemblance of the Prehellenic with the biblical types, the form of the names in the Theban and other series showing in many examples that they are not Hellenic or Semitic, but belonging to earlier epochs.

Under such investigation we shall find, as we want to find, earlier materials for Egyptian mythology, so far as it was not purely Egyptian, and we likewise obtain the means of better studying the language of the hieroglyphics. Hitherto, this has been dependent on another dead language, the Coptic, but in the Ude we have a living Egyptian, and of the earliest type, and in the Agaw languages we have elements for dealing with some other points of formation. Thus, we may carry out for Egyptians a comparative grammar.

The history of the Hebrew migrations involves probably a mixture and confusion of two or more examples of Cush, Havilah and Mitzraim interchanging Caucasia and Africa. The history of the previous Egyptian migration may refer to a first occupation of Caucasia, and then an advance into Africa, where this Egyptian race may have acquired a civilisation it did not

possess in Caucasia. Under such circumstances, although Herodotus would still find in Caucasia an Egyptian-speaking population, there might be no hieroglyphics, and no monuments of the types now so familiar to us.

It may be suggested that circumcision was derived by Egyptians and Syro-Arabians from earlier races, and propagated from a common centre, passing into Africa, and being ultimately better preserved in Arabia and Africa.

The fact that the Udes, Abkhass, and all other Caucasians are no longer black, needs little space in explanation. While the languages have in some cases resisted the invasions of the Georgians, Armenians, Greeks, Persians, and Turks, constant intermarriages with the invaders have replaced the aboriginal types, but not without bearing evidences of survival.

The study of the Ude language and population, as well as that of others in the Caucasus, is of great importance in all historical investigations, because it will greatly assist in laying better foundations for history. The language of the few hundreds who now speak Udish will, under the invasion of Turkish and Russian, in our time perhaps cease to live, and the collection of every fact, however small, however isolated, is valuable, because one fact may be the connecting joint or link of a chain of evidence otherwise incomplete.

DISCUSSION.

Dr. LEITNER was at a loss to understand why he should have the honour of being the first to be called upon to make a few remarks, unless it was an act of great courtesy and kindness extended to him, as he had just returned from Northern India with the results of, perhaps, interesting explorations. Among these, one of the most prominent was connected with the Khajunà language, spoken in Hunza. Dr. Clarke had caused some attention in India by his efforts to trace an analogy between that language and certain dialects spoken in the Caucasus and among perishing tribes there and elsewhere. If he succeeded, he would bring us back to a period far anterior to that assigned by popular, as well as even scientific, superstition to the Sanscritic form of language. With regard to the "Egyptian Colony and Language in the Caucasus," he had only been in the room for a few minutes, and had had no time to examine, even cursorily, the material on which Dr. Clarke's conjectures were based. He must, therefore, leave their discussion to the able men who were present, and who had heard the paper from the beginning.

The Rev. A. Löwy adverted to three of the subjects of which Dr. Hyde Clarke had treated in his lecture on "The Egyptian Colony in Colchis," viz. :

1. Circumcision.

2. The land of Havilah, and

3. The affinity which, it was argued, can be discovered to exist between the language of Egypt and that of Ude.

1. In regard to circumcision, Mr. Löwy stated that the ancient Hebrew records incidentally support the opinion that the practice of that rite was known to tribes that were held to be "brothers" of the Hebrews (namely the tribes notoriously belonging to the Semitic race), and also to nations that appear to have had a pedigree different from that of the Hebrews. The passage in Genesis xxxiv, verse 22, suggests the possibility that the mode of performing the operation of circumcision differed amongst various tribes or nations; for it is mentioned there, that every male of the subordinates of Hamor, the father of Shechem, should be circumcised "as they, the Hebrews, are circumcised."

2. Havilah, the Eldorado of the ancient Hebrews, is mentioned in the Jewish Scriptures as an appellative, with the definite article. As such, the meaning of the underlying actual or conjectural etymon was clearly understood by those whose vernacular language was Hebrew, and who connected it with the radical term (Hayil), which signi

fies "wealth," "abundance," "power," etc. In the early genealogies (Genesis x), Havilah is named as a brother of Sheba and Ophir, amongst the tribes inhabiting Southern Arabia.

3. To establish on a sound basis the connection of different languages of Asia and Africa, or of any other part of the world, the principles of each of the languages concerned in the comparison must be fully ascertained; and the laws by which transmutations are possible must be as clear as are those which are in operation in the Aryan stock of languages. The opportunity of recognising the alleged connection between the Egyptian and the Ude languages was insufficient to anyone who, like Mr. Löwy, had only seen one of the tables of a comparative vocabulary, drawn up by Dr. Clarke. Mr. Löwy had counted ten words, in which there was, more or less, a resemblance between Coptic and Ude words. As it would be contrary to rules of philology to hold that any merely casual resemblances prove the common origin of languages under examination, further proofs must be had before a safe opinion on the affinity of the two languages can be legitimately formed.

Mr. E. B. TYLOR remarked that the claim put forth by Mr. Hyde Clarke, to have discovered a connection between languages of the Caucasus and of Africa, to say nothing of South America, was one which would require very strong and precise evidence to substantiate it. He regretted that the tables of similarities on which the author relied had not been set up on the walls, for the meeting to judge of their extent and weight, and he pointed out, that in the only list circulated which had happened to come into his own hands, several comparisons of words were made in which there was no similarity at all. The meeting, he thought, would do well not to express any acquiescence in Mr. Clarke's views till they had fuller means of judging of the weight of the evidence.