stated, but that the attempt must be illusory because the conclusions are not determinable from the premisses. The economic laws do not follow from any simple rule about human desires, because they vary according to the varying constitution of human society; and any conclusion which you could obtain would be necessarily confined to the abstract man of whom the law is supposed to hold good. Every such method, therefore, if it could be successful, could only lead to conclusions about human desire in general, and could throw no light upon the special problems of political economy, which essentially depend upon varying industrial organisation. I will not, however, go further. You must either, I hold, limit Political Economy to the field of statistical inquiry, or admit that, as a part of sociology, it deals with questions altogether beyond the reach of mathematics. Like physiology, it is concerned with results capable of numerical statement. The number of beats of the pulse, or the number of degrees of temperature of the body, are important data in physiological problems. They may be counted, and may give rise to mathematically expressible formulæ. But the fact does not excuse us from considering the physical conditions of the organs which are in some way correlated with these observed phenomena; and, in the case of Political Economy, we have to do with the social structure, which is dependent upon forces altogether incapable of precise numerical estimates. That, at least, is my view; and I shall apply it to illustrate one remark, which must, I think, have often occurred to us. Political Economy, that is, often appears to have a negative rather than a positive value. It is exceedingly potent--so, at least, I think--in dispersing certain popular fallacies; but it fails when we regard it as a science which can give us positive concrete "laws". The general reason is, I should say, that although its first principles may be true descriptions of facts, and any denial of them, or any inconsistent applications of them, may lead us into error, they are yet far from sufficient descriptions. They omit some considerations which are relevant in any concrete case; and the facts which they describe are so complex that, even when we look at them consistently and follow the right clue, we cannot solve the complicated problems which occur. It may be worth while to insist a little upon this, and to apply it to one or two peculiar problems.