the old Priscianic form, of old and little-known dialects. But all these, with a vast amount of linguistic and grammatical lore, were scarcely scientific. A good many of them were rather inclined to believe Greek and Hebrew to be the parents of languages, and to consider Latin to be a derivative from Greek, Arabic an impure form of Hebrew, and Turkish and Persian both barbarous corruptions of Arabic. The comparative science of language, the methodical classification of dialects, is one of our own days: the name we require for it is Glossology, or Dialectology, the science of tongues or dialects: and one regrets that a word so inappropriate as Philology should have received the sanction of usage. No philosopher would dare, of course, to violate the rule of Bacon (de Aug. Sc., iii. 4): "Nobis decretum manet, antiquitatem comitari usque ad aras, atque vocabula antiqua retinere, quanquam sensum eorum et definitiones sæpius immutemus." But let us hope the "vocabulum " is not yet so "antiquum " as to be unchangeable. The German "Sprachkunde" is excellent, but "speech-cunning" would be uncouth to our ears, might perhaps mean Rhetoric, or the art of eloquence, and would be at variance with our rule (the rule of Linnæus) to employ no scientific names but those derived from the Greek. Perhaps "Dialectology" may eventually obtain favour. It will have the virtue (which "Philology has not) of really meaning what it stands for. Though "verba notionum tesseræ sunt," Bacon did not mean that the counter was to be stamped with the externals of another notio than the one it represented.

If we picture to ourselves a man with a keen ear and an observant mind, standing in some open spot in the great fair of Nijni Novgorod, we can imagine what a host of subjects for thought must be aroused and enter that mind, from the varied sounds which would strike that ear. The soft but sibilant Russ, the softer and less sibilant Servian, the harsher Bulgarian, the easy-flowing Osmanli, the rougher and more diversified Turkoman, Bashkir, and Mongol; the grunting Chinese, the guttural Arabic, the elegant and stately Persian, perhaps the strange Circassian, Georgian, Ossetic, the ear-breaking Pushtoo, mingled possibly with some sonorous tongue from the south of the Himalaya, and with the strongly accentuated dialects of Latins or Germans from the West, would meet in his sensorium with an apparently unmeaning tumult. And yet it would be clear, on reflection, that this was no tumult, nor yet unmeaning. Those varying sounds might all be observed to vary according to some law, and to recur at certain intervals; each set

of sounds would be found to have its peculiar character, distinguishing it from other sets of sounds; and the character and laws of variation of one set would be found to approximate more or less to those of some of the other sets, and to differ more or less notably from those of others. And it would soon occur to a thoughtful mind that those various sets of sounds might be grouped, and the groups subdivided with reference to the greater or less similarity of their character and laws. Such grouping would be a " Philology," or Dialectology. What we have fancied as presenting itself to the mind of our thinker at Novgorod, has occurred to the minds of men who have observed the similarities and differences of the various modes of communication by articulate sounds in use among mankind; and the result has been that science of classification of languages which we we term Comparative Philology.

Philologers have as yet definitely pointed out only certain great families of languages, which they distinguish from one another mainly by their grammatical characteristics.

1. The simply monosyllabic, in which one word of one syllable stands for one idea, and these words are never altered, but relation is expressed by their arrangement in order in the sentence. The type of these is the Chinese.

2. Those in which relation is expressed by attaching to the original root a number of monosyllabic or dissyllabic suffixes, the root remaining almost or entirely unchanged. These are termed agglutinative, and the family is usually named Turanian. The type of them is the Turkish.

3. Those which express relation by a system of prefixes and suffixes, joined to a root mostly monosyllabic, but variable in form. These are termed Hamitic, and their type is the Coptic. The family seems to extend through the whole of Africa; but as the great majority of these modern African tongues are entirely without literature, and none are written, their classification is by no means easy, nor has the task yet been carried very far.

4. Those which express relation by a system of suffixes almost entirely monosyllabic, and a very few prefixes, joined to a root normally dissyllabic, and very slightly variable. These are termed Shemitic, and their type is Arabic or Hebrew.

5. Those in which relation is expressed by variations in the middle or ending of a root primarily monosyllabic, but derivatively polysyllabic. These are called Aryan, and the type of the family, a very large and varied one, is Sanskrit.

6. To these we may add the family of languages spoken in the islands of the Pacific. They have not yet been regularly

classified; and some are of opinion that they may be considered as offshoots of the Malay, which is itself (they imagine) to be referred to the Aryan family. The peculiarity of these languages is that the words and their inflective particles are simple syllables, consisting of a consonant and vowel, or in some cases of a single vowel. They might be termed polysyllabic.

7. The languages of Northern America are characterized by the same colligation of syllables; but as the syllables are compound, and the whole system of colligation more complicated, some incline to group them with the Turanian or agglutinative, some to consider them a special family, the polysynthetic. We have here, then, seven families of human speech; or, to reduce them to the very lowest number, by classing the Polynesian with Aryan, the Shemitic with Hamitic, and the American with Turanian-at least four different forms of language.

But the clear statement of Scripture is that there was a time when "all the earth was one lip, one set-of-words " (I translate Gen. xi. 1, literally). Their vocabulary and their pronunciation were the same.

Here the opponents of Scripture join issue. They tell us that, do what we will, we cannot avoid the conclusion that the various families of languages, be they seven or four, or any ultimate number, exhibit such specific differences that they cannot have been developed from one original; that, in fact, the diversity of human speech is as good and convincing an argument in favour of the polygenist hypothesis as the diversity of human physiology.

But this is rather a violent assumption. What proof is there that the differences in human languages, great as they are now, are so essential that they may not be explained by the disturbing and disorganizing causes which are at work even amongst ourselves, and are productive of speedy effects where there is no written literature to give fixity to the vocabulary and grammatical forms? Granted that Chinese and Sanskrit, Siamese and Gaelic, Finnish and Kafir, are SO utterly and entirely dissimilar now, that we can scarcely imagine the human being who has learnt the one acquiring the power of using the other, that dissimilarity is not other in kind, it is only greater in degree, than the difference between a page of the Saxon Chronicle and a page of the Times; or to use a still better illustration, than that between an upnekhat of the Zend-Avesta and a division of the Shah-Nameh, or a proclamation by the present Shah of Persia, between the Dutch Bible and Ulfilas.

The disturbing forces which act upon language are in the main the following;-I postpone, of course designedly, that supernatural disturbing force which we of this Institute believe to have been injected into humanity in the plain of Babel; and to have been, temporarily and in part, lulled in the early days of Christianity after the great day of Pentecost:

1. National or tribal peculiarities. Those anatomical or physiological peculiarities which constitute the differences between races of men are not without effect upon their speech. The inhabitants of a southern climate, and of a richly fertile territory, naturally fall, after a generation or two, into slothful unenergetic habits. They speak lazily; they shrink from the difficulty of hard consonantal pronunciation, and complicated inflexion. Compare the Polynesian tongues with every other family; or, to come to differences in the same family, contrast the soft Italian with the harder Rumonsch of the mountains; Servian with Polish; Bengali with Mahratta,-nay, the English of Aberdeen with the English of Exeter. Again, a peculiar conformation of the organs of speech, produced by some external cause, climatic or otherwise, would soon eliminate some sounds, and introduce others; and thus, if I may so express it, the tuning of the national ear would take a particular direction, and the pronunciation and vocalization of the language would have a tendency to alter towards one class of sounds, and away from another class. As an instance of this "tuning" as I have called it, I may allege the aversion of the Italian ear to a number of consonants in juxtaposition. Such a sentence as "with great strength and speed" is positively terrible to a nation which cannot say il but lo sbaglio, and turns Xerxes into Serse. Another example is the rigid rule of harmonizing sounds in Turkish, according to which a flat suffix must follow a flat root, and a sharp suffix a sharp root: e. g. (ye-mec, to eat); but F. (yu-mak, to wash). Another perhaps is the rejection, as offensive and barbarous, of the clicks which are so prominent in the language of the Bosjesmans and some few other African tribes; not only are they found in no other family of tongues, but the higher Kafirs, as the Sechuana, never employ them.

Further, habits of mutilation or distortion, not uncommon among barbarous tribes, must exercise a great influence in modifying language. Dental sounds and sibilants must be considerably altered, if not utterly lost, among those who file away or strike out the front teeth. Distortion of the lips, too, must interfere with the articulation of labials. So also among the imperfectly civilized, the habits of mutual suspicion and

dread lead to a plan of speaking with as little apparent movement of the face as possible; hence labials and fine distinctions in vowels disappear, and gutturals, with slight modifications of the "ur-vocale" (Sanskrit) take their place in the development.

2. Not only national peculiarities, but those of individuals, influence the language of a tribe. A natural defect in the articulation of a powerful chieftain would lead his followers, out of respect, to imitate that very defect, or at least to conceal their possession of superior powers of speech.. Even amongst ourselves we can often observe a tendency to affect some peculiarity in the enunciation or mode of expression of a leading man; his very phrases are caught up and incorporated into the language of his admirers. In the days of unwritten language such imitation must have had a very decided and permanent effect upon the speech of a tribe.

3. A fertile source of variations in dialect is the tendency to imitate the imperfect pronunciation of children, and to clip and alter words in order to adapt them to their untrained organs. Cases of this kind are familiar to ourselves. There is scarcely a family in whose domestic language some eccentric phrase or mis-pronunciation has not become current, derived from the prattle of some one of its youthful members. Such disturbances as these are of course counteracted by the comparative fixedness of a written language: the family argot is confined within the circle in which it was produced. But in earlier days, without this impediment to change, as in illiterate tribes at this day, the mimicry of children was doubtless a powerful disturbing force, affecting not only the forms, but the grammatical inflexions of words, and their collocation in

sentences.

4. Superstition in less civilized tribes, and, to a slight extent, social rules in more civilized communities, affect the language. Many words and phrases which were usual in this country two centuries ago have become offensive, quaint or ridiculous, and as such are practically banished from our normal literary tongue, though they linger in our provincial dialects. The verbal inflexion in th (hath, goeth, &c.) is now quite lost in classical English, though it was current a century ago, and common at double that distance of time. Now, if an inflexion can be lost in this manner out of a written language in whose literary remains it is of continual occurrence, it is plain that under circumstances of less restraint the process of alteration would go on more rapidly; and two portions of the same tribe, separated from one another by a range of mountains or an arid plain, might find, after half a century without