Hence, the Paropamisans may safely be considered as a population of a receding frontier, the encroachment upon their area having been Afghan. With these the Asiatic populations end.
If we now look back upon the ground that has been gone over, we shall find that the evidence of the human family having originated in one particular spot, and having diffused itself from thence to the very extremities of the earth, is by no means absolute and conclusive. Still less is it certain that that particular spot has been ascertained. The present writer _believes_ that it was somewhere in intratropical Asia, and that it was _the single locality of a single pair_--without, however, professing to have proved it. Even this centre is only _hypothetical_--near, indeed, to the point which he looks upon as the starting-place of the human migration, but by no means identical with it. The Basks and Albanians he does not pretend to have affiliated; but he does not, for this reason, absolutely isolate them. They have too many _miscellaneous_ affinities to allow them to stand wholly alone.
In the way of physical conformation, the Hottentot presents the _maximum_ of peculiarities. The speech, however, of the latter is simply African; whilst, in form and colour, the Basks and Albanians are European. A fly is a fly even when we wonder how it came into the amber; and men belong to humanity even when their origin is a mystery. This gives us a composition of difficulties, and it is by taking this and similar phænomena into account, that the higher problems in ethnology must be worked. Nothing short of a clear and comprehensive view of the extent to which points of difference in one department are compensated by points of likeness in another, will give us even a philosophical hypothesis; all _partial_ argument from partial points of disagreement being as unscientific as a similar overvaluation of resemblances.
As for the detail of the chief difficulties, the writer believes that he, unwillingly and with great deference, differs from the best authorities, in making so little of the transition from America to Asia, and so much of that between Europe and Asia. The conviction that the Semitic tongues are simply African, and that all the theories suggested by the term _Indo-European_ must be either abandoned or modified, is the chief element of his reasoning upon this point--reasoning far too elaborate for a small work like the present. He also believes that the languages of Kafferistan, the Dardoh country, and north-eastern Afghanistan, are transitional to the monosyllabic tongues and those of Persia--in other words, that the modern Persian is much more monosyllabic than is generally supposed. Yet even this leaves a break. How far the most _western_ tongue of this class can be connected with those of Europe, and how far the most _south_-western one has Semitic affinities are questions yet to examine--questions beset with difficulties. However, as the skeleton of system he believes the present work to be true as far as it goes, and at the same time convenient for the investigator. That there is much in all existing classifications which requires to be unlearnt is certain. Lest any one think this a presumptuous saying, let him consider the new and unsettled state of the science, and the small number of the labourers as compared with the extent of the field.