bold and wise to go even beyond this. There is very little that deserves to be called a national literature in the modern dialects of the Hindus. The sacred, legal, and poetical literature of India is either Arabic, Persian, or Sanskrit. Little has grown up since, in the spoken languages of the day. Now it would be hopeless, should it ever be attempted, to eradicate the spoken dialects of India, and to supplant them by Persian or English. In a country so little concentrated, so thinly governed, so slightly educated as India, we cannot even touch what we wish to eradicate. If India were laid open by highroads, reduced by railways, and swamped by officials, such an attempt might be conceivable, though, as to anything like success, a trip through Wales, and a glance at the history of Wales and England, would be a sufficient answer. But what might be done in India, perhaps even now, would be to supplant the various native alphabets by Roman letters. How few people in India can write! and those are just the men who are open to Government influence. If the Roman alphabet were taught in the village schools which the Government has much encouraged of late, particularly in the north-western provinces, if all official documents, whatever languages they might be in, had to be transcribed into Roman letters before they had legal value, if the Government would issue all laws and proclamations transcribed in Roman characters, and if the missionaries would do the same with their translations of the Bible and other publications in any of the dialects of India, I think we might live to see one alphabet used from the "snows" to Ceylon.
In countries where there is a growing and living national literature, the experiment to supplant their national alphabet would probably fail; but India is the very country where it might be tried with a fair chance of success.
Let us see, then, how our physiological Missionary alphabet could be applied to languages which have not only an alphabet of their own, but also an established system of orthography.
We have here to admit two new principles:—
First, that in transliterating written languages, every letter, however much its pronunciation may vary, should always be represented by the same Roman type, and that every Roman type should always represent the same foreign letter, whatever its phonetic value may be in different combinations.
