merely in the example under view. In a complete theory of reasoning this topic would claim fuller investigation, but in Logic, which is understood to deal with communicable reasoning, and has therefore to deal with formulæ patent to all who are present, it does not further concern us.

III. The next question we have to consider is the nature of a language regarded in itself, that is, the particular kind of medium of communication which is to be selected. We take it for granted that the primary function of language is to communicate thought from one person to another; or, in so far as they are able to make use of it, from one sensible and intelligent being to another. For this purpose some system of sensible signs is needed, some medium which can be appealed to by more than one person at the same time. Granted that,—as indicated just above,―any one of us could carry on a process of solitary thought by a succession of images, yet if we are to communicate our thought to others we must know what images others are thus entertaining, and we must somehow succeed in exciting those that we wish. That is, we must be able to step outside our own personality sufficiently to know what is going on inside the minds of others. This at once raises the question as to how many different kinds of language there may be which would fulfil this general object; and when this has been settled, whether any one of them will answer the purpose better than another. At a first glance this seems to open out to us a wide array of possible alternatives. Language, we say, is to consist of a system of sensible signs. Now we have five distinct senses, in the ordinary signification of that term, any one of which may serve as an inlet for the sign selected. And as regards the modes open to us for appealing to the senses the choice is practically unlimited, as almost any action we perform will affect some sensible organ of those who are in our neighbourhood. We will take these two different modes of approaching the subject in turn, viz. the sense appealed to and the mode selected for appealing to it. The distinction is important though it is often overlooked: there is, for instance, as we shall presently remind the reader, no absolutely necessary reason why a language of sounds (perceived through the ear) should likewise be a vocal language (produced through the throat and lips).

(1) Look then first to the sense through which the symbols

constituting the language are to be recognized; laying aside for the present all consideration of the special mode to be adopted in exciting the sense, that is, in creating the symbol.

As regards two of these inlets of knowledge we may make a rather summary rejection. Taste and smell will not adequately answer our purpose. This is not owing, it must be remarked, to any inherent incapacity on their part to excite definite notions by association. Quite the contrary. The adhesive power between the sense-element and the notion is particularly strong in the case of one of these, namely, smell. Everyone must have noticed how the perception of definite smells instantly and powerfully raises in the mind notions of things connected in any way with them. And this takes place without the slightest previous cultivation of the sense for this purpose, or intentional employment of it in this way, which seems to intimate that in this direction we have one at anyrate of the conditions for a good language. Moreover the range and diversity of distinct perceptions here is very considerable; it is quite a question whether the number of recognizably distinct tastes and smells is not comparable with that of sounds.

The objections to these senses as inlets of speech-symbols seem to be mainly the following. For one thing they are very liable to variation from time to time, and even to almost total loss. A fit of indigestion or a bad cold would have the effect of rendering us what we may call by analogy blind or deaf. Again, these senses are apt to interfere mutually with each other when impressions occur successively. There is something of this sort even in the case of sight; what are known as "after images" being a modification of the present sense-datum by those which had gone before. But for all practical purposes visible images can succeed each other rapidly without suffering any disturbance in their normal condition, whereas any one taste or smell-perception (particularly the former) may be largely modified by the effects which the preceding ones have left behind them. Above all, we seem to possess an extremely small power of combining several elementary impressions into practical simultaneity, so as to obtain a total impression of which the parts shall still be distinguishable. In the case of sights and sounds we can carry this a long way. For instance, a whole word, in writing or printing, can be gathered up into a unit without the

parts merging indistinguishably into each other. We may only visualize clearly one of the letters at a time, but we can practically retain the whole group before us so that they seem to be present simultaneously. This adds enormously to the range of variety of symbols which can be produced, for a few simple signs can be readily built up into a multitude of more complex ones. In the case of smells and tastes we seem to possess scarcely any power of this kind; for I suppose that the pleasure felt by the gourmet in the infusion of a well-adjusted sauce, or the cunning composition of a dish, arises either from mere harmony or contrast in succession, or from the production of some new and apparently single flavour. These considerations have an important bearing when we reflect how large is the part which is played in all our processes of thought by the closely allied operations of analysis and synthesis. It then becomes very important that our symbols of language and of thought should be themselves capable of readily falling into a complex form which it is yet possible to resolve easily into parts.

There then remain the three senses of sight, hearing, and touch,—including under this last our powers of perceiving muscular resistance and temperature, which are closely connected with touch proper. All these three, but especially the two former, possess in a high degree the main requirements we demand. Still regarding them merely from the point of view of inlets of perception, we must put to their credit the following prominent merits. They are but slightly affected by bodily or constitutional conditions and defects: they offer us a wide range of distinctly perceptible elements: they are sufficiently persistent or recoverable in their integrity for us to be able to grasp up a number of their constituent elements into what is for all practical purposes a simultaneous group: and any one of these perceptions exercises but little disturbing influence upon those in its neighbourhood. So far there would not seem to be much reason to prefer one of these three to another, and in fact all of them are employed for the purpose:-sounds, by all intelligent beings; sights, by all the cultivated and literary races and classes; and touches, by all who are deprived of the other two senses or who wish to supplement their shortcomings.

(2) But this is only one side of the question. Besides the mode of taking in an impression there is also to be considered

the mode of exciting it. On this side we have before us not five alternatives merely, but an altogether indefinite number. Our possible means of making noises, producing visible images, tastes, and so forth, are endless. Not only do our own limbs, and vocal and other organs, supply means which are ever at hand, but there are also innumerable artificial appliances to which we might resort for aid. It is the enormous superiority of some of these over others that is the really determining factor in the choice of the particular kind of language to which we are all accustomed.

As regards two of our senses,-taste and smell,—we found that they stood at a considerable disadvantage regarded solely as inlets of impressions. When we proceed to discuss the comparative opportunities afforded us of exciting the requisite impressions, the hopelessness of any expedient of this kind is obvious at once. To produce the same taste simultaneously in every palate in a large assemblage is in itself a matter of some trouble and outlay. But if we wanted to construct a language on such a basis we should have to invent some mode of rapidly changing these tastes from moment to moment. The stammerers' dinner recorded in the Spectator (No. 371), at which one gentleman took a quarter of an hour in communicating his conviction of the excellence of the ducks, whilst another expended as much time in expressing his assent to this statement, might certainly have been better carried out through any other sense; for they probably did not take so long to eat the food as to describe it afterwards. A suitable exchange of dishes, according to some conventional mode of interpretation, would have served to greatly abridge this dialogue; but then it must be admitted that the gentlemen in question, by choosing to resort to speech, adopted a mode of communication for which they were unusually badly adapted.

Smells stand on a slightly better footing. It would be quite a question whether something might not be done in the way of devising a sort of rudimentary language on this basis for the benefit of any unfortunate persons who were deprived of every other inlet but this. A succession of scents might spell out an alphabet, and a combination of them might even form a nasal image which should be simultaneously perceived by a small company, and raise the same notion in the minds of them

all. But fortunately we have incomparably better means than this available.

As regards the choice between the other senses, we shall best see the grounds of preference for one or another of them by noticing in turn the main requirements demanded in a successful language.

(i) The first of these requirements seems to be that the apparatus for exciting the requisite sensations in others should be always readily available. In this respect nature has left us but very little option. Our main resource, as every one knows, is sound: in fact, this is the only universal, or even approximately general, language. The reason for this almost exclusive preference clearly is the fact of our carrying about with us such an admirable apparatus for producing sounds at will as that which we call our vocal organs. How much of their present excellence they owe to gradual training in the individual, and to ages of exercise and inheritance in the race, does not now concern us. There can be little doubt however that, as far back as we can go, these particular organs must have enjoyed a preeminence for the purpose in question which makes any comparison with other organs out of the question. About the only other methods in which we well could make noises, without resort to any appliances but our own, would be by snapping our fingers, slapping our sides, and so on; as indeed sometimes is done now by way of emphasis. Crickets, as far as we know, have a sort of very rudimentary mode of communication by making noises through scratching their sides. And had a race of intelligent beings gradually become differentiated through this line of descent, they might not impossibly have ultimately developed some highly complex kind of stringed instrument about their own persons, as we may be supposed to have developed our present wind instruments inside our throats. Evolutionary forces care but little how far back they have to start, provided we give them time enough and suitably varied circumstances.

Our available resources for making visible signs, that is, for appealing to the sense of sight in others, are not quite up to the standard attained in the case of sounds. Much can be effected, of course, with the hands and arms in this way; but in the savage state men have something else to do with these organs,