with. For the former no reason can be given. For example, "afeard" is a more logical form than "afraid" because of its evident connection with fear, but it is no longer considered fit company for refined society. The history of "astun," "astony," "stun," "astonish," affords another instructive example. "Climbed" and "heated" have taken the place of "clomb" and "het," although they are longer and more expressive. Generally speaking those words that are the most used are the most irregular. Our verb of existence—am, was, been—is made up of three different stems. Our grammatical auxiliaries are very defective; the missing parts have to be supplied in various ways. We say "I must" for the present, but we can not say "I musted" for the preterit, nor is there such a verb in English as "to must" although it is found in the Anglo-Saxon. "Might" is usually classed as the preterit of "may," but in many of its uses it is not. In all the languages of the Aryan stock, and in their descendants, we find the same lack of parts and the same alien substitutions. The changes that have taken place within the historic period are just as difficult, in fact just as impossible, to account for as the earlier ones. If languages were constructed according to any system, or even according to the most elementary principles of common sense, they would differ widely from their present status. It may be said in passing that some of the languages of the Turanian stock, notably the Turkish, are to a considerable extent symmetrically built. It used to be said that many words have been modified in obedience to the general law that tends to ease of utterance; but this explanation is no longer accepted. If such a law was ever operative an inexplicable break in the continuity of the human psyche must have taken place at some remote period in the past. Such a break would be at variance with the wellestablished course of development. It seems probable that human speech originated at three or more points on the surface of the earth. As long as these primitive tongues were left to develop according to their innate laws the process was consistent, if not logical. But when two or more of these original stocks came into conflict, each party trying in its clumsy way to acquire the speech of the other, confusion set in. If ten crassly illiterate Frenchmen and ten equally illiterate Hungarians were placed together where they would be compelled to communicate with each other we may be sure that in two or three generations a language would be produced differing widely from either parent. We have practical examples in the mixture of Norman French and AngloSaxon, of Iberian with Latin, and elsewhere, although these instances are not precisely such as I have supposed above.
As the Oxford dictionary is about four sevenths completed, the entire work will include more than 350,000 words. It is claimed by the publishers of the Standard Dictionary that the latest edition
