into germs, plants, flowers, and fruits in the various languages of the Aryan race.

After having settled this historical outline of the growth of our family of speech, the Aryan, we take any word, or a hundred, or a thousand words, and analyse them, or take them to pieces. That words. can be taken to pieces, every grammar teaches us. The Sanskrit name for grammar is Vyâkarana, which means 'taking to pieces.' This process, however, of taking them to pieces scientifically and correctly, dissecting limb from limb, is often as difficult and laborious as any anatomical preparation.

Acutus.

Well, let us take quite a modern word-the American cute, sharp. We all know that cute is only a shortening of acute, and that acute is the Latin acutus, sharp. In acutus, again, we easily recognise the frequent derivative tus, as in cornutus, horned, from cornu, horn. This leaves us acu, as in acu-s, a needle. In this word the u can again be separated, for we know that it is a very common derivative, in such words as pec-u, cattle, Sanskrit pasú, from PAS, to tether; or tanú, thin, Greek Tavú-s, Latin tenu-i-s, from TAN, to stretch. Thus we arrive in the end at AK, and here our analysis must stop, for if we were to divide AK into A and K, we should get, as even Plato knew (Theaetetus, 205), mere letters, and no longer significant sounds or syllables. Now what is this AK? We call it a root, which is, of course, a metaphor only. What we really mean by a root is the residuum of our analysis, and a residuum which itself resists all further analysis. But what is important is that these roots represent not a mere theoretic

postulate, but a fact, an historical fact, and, at the same time, an ultimate fact.

With these ultimate facts—that is, with a limited number of predicative syllables, to which every word in any of the Aryan languages can be traced back, or, as we may also express it, from which every word in these languages can be derived-the historical school of comparative philology is satisfied, at least to a certain extent; but it has also to account for certain pronouns and adverbs and prepositions, which are not derived from predicative, but from demonstrative roots, and which have supplied, at the same time, many of those derivative elements, like tus in acu-tus, which we generally call suffixes or terminations.

After this analysis is finished, the historical student has done his work. AK, he says, conveys the concept of sharp, sharpness, being sharp or pointed. How it came to do that we cannot tell, or, at least, we cannot find out by historical analysis. If we like to guess on the subject, Plato has shown us how to do it, and no one is likely to do it more ingeniously than he. But that it did convey that concept, we can prove by words derived from AK in Sanskrit, Greek, Latin, in Celtic, Slavonic, and Teutonic speech. For instance: Sanskrit âsu, quick (originally sharp), Greek wκús, Lat. oc-ior, Lat. ac-er, eager, acus, acuo, acies, acumen; Greek aκμń, the highest point, A.-S. acg; also to egg on; akov, a javelin, acidus, sharp, bitter, ague, a sharp fever, ear of corn, Old High German ahir, Gothic ahs, Lat. acus, aceris, husk of grain, and many more.

Theoretic School.

Let us now look at the Theoretical School and its treatment of language. How could language arise?

it says; and it answers, Why, we see it every day. We have only to watch a child, and we shall see that a child utters certain sounds of pain and joy, and very soon after imitates the sounds which it hears. It says Ah! when it is surprised or pleased; it says Bah! when it sees a lamb, Bow-wow! when it sees a dog; and it soon says See-saw, when it swings its doll. Language, we are told, could not arise in any other way; so that involuntary interjections and imitations must be considered as the ultimate, or rather the primary facts of language, while their transition into real words is, we are assured, a mere question of time.

This theory, or rather these three theories, which have been called the Pooh-pooh, Bow-wow, and Yo-heo theories, are said to be easily confirmed by a number of words in all languages, which still exhibit most clearly the signs of such an origin; and still further, by the fact that these supposed rudiments of human speech exist, even at an earlier stage, in the development of animal life, namely, in the sounds uttered by dogs, parrots, and other animals; though, curiously enough, far more fully and frequently by our most distant ancestors, the birds, than by those who claim to be our nearest relatives, the apes.

It is not surprising, therefore, that all who believe in a possible transition from an ape to a man should gladly have embraced this theory of the origin of language.

The only misfortune is that such a theory, though it easily explains utterances which really require no explanation at all, such as bow-wow and cuckoo ; pooh-pooh and fie! yo-heo and see-saw, or even words

such as crashing, cracking, creaking, crunching, scrunching, leaves us entirely in the lurch as soon as we come to deal with real words-I mean words expressive of general concepts, such as man, tree, name, law-in fact, nine-tenths of our dictionary.

The Theoretical School has certainly one great advantage. It goes to the very bottom of the question, and explains the very origin of language, as it took place in the nursery of the first Pithecanthropos or Anthropopithecos, and it explains it in so simple a way that every child can understand it. If a child. can say Bow-wow, what difference is there between that and saying Dog? If a child can say Fie, why should it not say 'I disapprove'? If a child says Dingdong, why should it not say Bell? All these, we are told, are differences of degree only, whatever that may mean, and with a sufficient allowance of time, there is nothing that will not become anything.

The Historical School cannot match such performances. When by a most laborious analysis it has reduced one language, or one family of languages, to its constituent elements, it cannot claim to have accounted for the origin of all language, but only of one or two or three families of human speech. When it has placed before us the roots of one language, or one family of languages, it has come to the end of its work. It can do no more than leave these roots as ultimate facts, though between these roots and the first friendly grunts that passed between anthropopithecos and pithecanthropos, there may be millions of millions of years.

Then why not adopt the Bow-wow, the Pooh-pooh, and the Yo-heo theories, which explain everything so

easily and so completely? For the simple reason that real language, when we trace it back to its real constituent elements, shows no trace whatever of these mere imitations of so-called natural sounds. They exist not as part and parcel of real language, but simply by the side of it. Even admitting the possibility that they might have grown into some kind of language, the fact remains that they have not done so1. What we call roots do not only show no outward similarity with these natural sounds,-that would be the smaller difficulty, but they are totally different in nature; and this is the point which so few anthropologists seem able to see. These roots are not simply perceptual, like all Bow-wow, Pooh-pooh, and Yo-heo utterances; but they are conceptual in character, as the elements of conceptual language ought to be, if they are to help us to explain what has to be explained, namely, conceptual speech.

Brinton on Palaeolithic Language.

This has evidently been perceived by Dr. Brinton, now Professor of American Linguistics in the University of Pennsylvania. He knows that interjections and all the rest will not grow into real language. But he thinks that the American languages will help us to get behind the scene, and he has drawn up a picture of what, following their guidance, he imagines the language of Palaeolithic Man to have been 2. 'It was far more rudimentary,' he writes, than any

1 On the possibility of such a transition, see Science of Thought. pp. 309-315, a chapter for which I have been much blamed by scholars, while anthropologists have construed a limited concession into a complete surrender.

2 The Language of Palaeolithic Man, by Daniel G. Brinton, M.D. Read before the American Philosophical Society, October 5, 1888.