HKB
a transitional dialect before they settled into the forms now in use, the philologist might bring forward proofs of the great numbers both of lost and living forms of speech.
Strabo tells us that in his time, in the Caucasus alone (a chain of mountains not longer than the Alps, and much narrower), there were spoken at least seventy languages. At the present period the number, it is said, would be still greater if all the distinct dialects of those mountains were reckoned. Several of these Caucasian tongues admit of no comparison with any known living or lost Asiatic or European language. Others which are not peculiar are obsolete forms of known languages, such as the Georgian, Mongolian, Persian, Arabic, and Tartarian. It seems that as often as conquering hordes swept over that part of Asia, always coming from the north and east, they drove before them the inhabitants of the plains, who took refuge in some of the retired valleys and high mountain fastnesses, where they maintained their independence, as do the Circassians in our time in spite of the power of Russia.
In the Himalayan Mountains, from Assam to its extreme north-western limit, and generally in the more hilly parts of British India, the diversity of languages is surprisingly great, impeding the advance of civilisation and the labours of the missionary. In South America and Mexico, Alexander Humboldt reckoned the distinct tongues by hundreds, and those of Africa are said to be equally numerous. Even in China, some eighteen provincial dialects prevail, almost all deviating so much from others that the speakers are not mutually intelligible, and besides these there are other distinct forms of speech in the mountains of the same empire.
The philologist might next proceed to point out that the geographical relations of living and dead languages favour the hypothesis of the living ones having been derived from the extinct, in spite of our inability, in most instances, to adduce documentary evidence of the fact or to discover monuments of all the intermediate and transitional dialects which must have existed. Thus he would observe that the modern Romance languages are spoken exactly where the ancient Romans once lived or ruled, and the Greek of our days where the older classical Greek was formerly spoken. Exceptions to this rule might be detected, but they would be explicable by reference to colonisation and conquest.
As to the many and wide gaps sometimes encountered between the dead and living languages, we must remember that it is not part of the plan of any people to preserve memorials of their forms of speech expressly for the edification of posterity. Their manuscripts and inscriptions serve some present purpose, are occasional and imperfect from the first, and are rendered more fragmentary in the course of time, some being intentionally destroyed, others lost by the decay of the perishable materials on which they are written; so that to question the theory of all known languages being derivative on the ground that we can rarely trace a passage from the ancient to the modern through all the dialects which must have flourished one after the other in the intermediate ages, implies a want of reflection on the laws which govern the recording as well as the obliterating processes.