tion of the errors which have marked the course of economic research. I hold, therefore, that it is of the utmost importance, not only in Political Economy, but in all social investigation, that the terms of our nomenclature should, as far as possible, serve as constant reminders of the nature of the concrete objects which they are employed to denote; and that for this purpose, to borrow Mr. Mill's language, "as much meaning as possible should be thrown into the formation" of our economic terms, "the aids of derivation and analogy being employed to keep alive a consciousness of all that is signified by them.'

It will serve to throw light at once on the resources at the disposal of the economist in this respect, and also on the special difficulties under which Political Economy labours in the matter of definition, if we advert for a moment to the case of the physical science which offers the most perfect example of a nomenclature framed on the principle we have now in view. This is chemistry, in which the nomenclature is at once significant and technical-significant, inasmuch as its terms are composed of elements taken either from existing or from the ancient languages which carry their original meaning into their new occupation; and technical, inasmuch as in their actual form they are only employed as members of a scientific nomenclature, Such words as oxygen, hydrogen, carbonate of lime, peroxide of iron, are all full of meaning, but are never employed except to express certain known chemical elements or combinations. From this union of the two qualities of significance

and technicality in its nomenclature an immense advantage results for chemical science; since its terms. have in consequence the power of calling up with great distinctness the concrete objects they are intended to denote; while, having been constructed for the special purpose of designating those objects, and never being employed in common speech, they are free from all associations which could confuse or mislead either those who employ or those who hear them. The point, then, to be considered is how far it is possible to construct for Political Economy a nomenclature which shall fulfil the same ends as nomenclature in chemistry. It appears to me that a certain approximation towards this result is feasible, but only an approximation, and that, after all is done, the technical language of Political Economy must ever fall vastly short of the perfection attained by terminology in chemical science. In coming to this conclusion I assume it as settled that the technical terms of Political Economy are to be taken from popular language, and this, not merely as regards their elements, as is done in chemistry, but, so to speak, bodily in their complete forms. Whether it would, at any time, have been possible to have constructed an economic nomenclature on the plan adopted in chemistry is perhaps scarcely worth considering. The science has, in fact, been developed through the instrumentality of popular language. It is through this medium that the ideas of all its greatest thinkers have been put forth; it is in this clothing that the world is familiar with them; and it is, therefore, now

palpably too late, even if there were no other restraining consideration, to think of recasting its doctrines in other forms. Such words as production, distribution, exchange, value, cost, labour, abstinence, capital, profit, interest, wages, must now for good or for evil remain portions of economic nomenclature; and these have all been drawn in their actual forms from the vernacular, and are in constant use in popular speech. With regard to such words they are capable enough of fulfilling the first of the two functions fulfilled by nomenclature in chemistry—of calling up, that is to say-always supposing them to be used with deliberation-concrete facts and objects with sufficient vividness. The hitch occurs in their inaptitude for the second of the two purposes required of them, for bringing to the mind the exact facts and objects, neither more nor fewer, which we desire to indicate.

For the position of things is this:-The economist finds it necessary, for the reasons which have been stated above, to arrange the phenomena of wealth in classes on a certain principle-that principle being in fact the convenience of his own. investigations; and he has to find names for the classes thus constituted in the terms of popular language. But popular language has not been framed to suit the convenience of economic speculation, but with quite other views. Its distinctions and classifications do not always or generally coincide with those which are most important for the elucidation of the economy of wealth; and even where

L

this correspondence is tolerably close, a term in constant use in ordinary speech inevitably gathers round it a vague aroma of association, sure to suggest in particular contexts ideas which have no proper connection with the purposes of scientific research, and which therefore cannot but act as hindrances to the reasoning process. That precision of meaning, accordingly, which is so conspicuous in the nomenclature of chemistry, and in general of the physical sciences, is unattainable in Political Economy. Its nomenclature satisfies indeed the condition of having plenty of meaning. With even greater vividness than the nomenclature of chemistry it is capable of calling up the concrete things denoted by its terms; but for this advantage it pays the heavy price of loss of precision -of vagueness and uncertainty as to the proper limitation to be given to its most important words. The remedy, so far as remedy is possible, seems to be twofold:-first, to keep our definitions of economic terms as close to the usages of common speech as the requirements of correct classification will allow. Terms must indeed, now and then, be strained to express meanings and to suffer limitations which in ordinary discourse they do not express or bear, since otherwise the ends of classification would be sacrificed; and it is, therefore, no conclusive objection to an economic definition that it does not accurately coincide with popular use. But it should, nevertheless, be fully recognised that such deviations constitute a demerit in definition, and may become a serious one. The second remedy against the evil

is clearness and distinctness of definition wherever terms of importance are employed; care being taken, where the economic sense differs from the popular one, to bring into as strong relief as possible the points of difference; with which precaution the practice may be usefully combined of throwing in a caveat from time to time, where the context would be in danger of suggesting the popular rather than the scientific sense.

§ 5. We may now sum up the general results of the foregoing discussion :-

1. The first requisite of a good definition in Political Economy is that it should mark those distinctions in facts and objects which it is important to mark with a view to the elucidation of the phenomena of wealth; and our nomenclature will be good or bad, helpful or obstructive, according as it coincides with such real and pertinent distinctions, or sets up others which are arbitrary, fanciful, or irrelevant.

2. So far as is consistent with satisfying the foregoing condition, economic terms should be used as nearly as possible in their popular sense; though, as strict adherence to popular usage is not compatible with fulfilling the requirements of sound classification, the mere circumstance of deviation from popular usage is no conclusive objection to

an economic definition.

3. It is no valid objection to an economic definition that the attribute on which it turns is found to exhibit degrees in its concrete embodi