the proper business of ethics and jurisprudence. These sciences expound the nature of imputation or guilt, of moral responsibility, of intention, and negligence; and thus establish the grounds which determine the moral quality of an act, by circumstances independent of the declarations of the agent himself.

That part of human action which is mental, which is distinct from the mere bodily operation of the will, and which invests the act with its moral quality, is concealed from our senses; and as it is never the subject of direct observation, is always inferential. It is not the province of politics to define the grounds upon which these inferences are made, or to assign the marks by which the moral quality of an act is recognised. The analysis of this indirect process lies at the very foundation of the moral sciences; the solution is assumed by political science as known, and is not contained in it. In politics, however, as in ethics, the external act, taken by itself, is often ambiguous; it must be construed, or interpreted, by its circumstances. The act itself, and the circumstances, are matters of observation, and by these its quality must be determined. The mental state of the persons principally concerned in it can never be known to us directly. The most direct guide to their internal thoughts and feelings is their own declarations respecting them; and these declarations cannot always be trusted. In fact, they are not unfrequently made with the express purpose of misleading.

To illustrate what has been said, let us take some wellknown historical fact; for example, the execution of Charles the First. So much of this event as consists in outward acts-the imprisonment of the king, his trial and condemnation to death, and his decapitation upon a scaffold at Whitehall, on the 30th January, 1649, is vouched by undoubted testimony, and is as certain as that Herculaneum was destroyed by an eruption of Vesuvius, and Lisbon by an earthquake. But what were the reasons which prompted this act; what were the precise motives which induced Cromwell and the other leaders of the army to adopt this measure, can be known only by inference from their own conduct and demeanour, from the conduct of the

king, from their statements, and from the circumstances in which they were placed. (6)

Now in the physical sciences there is nothing which, in strictness, resembles moral action. All the phenomena succeed one another according to invariable laws of sequence—the ideas of volition, freedom of action, moral responsibility, punishment, government, self-determined progress, and self-created improvement, are inapplicable to matter, and to the realm of outward nature. All physical phenomena, in their inception, connexion, and cessation, are the subjects of observation by means of the senses. There is no mental state, imperceptible by the senses, which gives a form and colour to the material observed appear

ances.

The nearest approach to an analogy with this peculiarity of the moral sciences is to be found in the names of certain material agents, which we suppose to exist, though we know them only in their effects-such as heat, light, and electricity; even life and matter properly belong to this class. Natural theology, likewise, so far as it is to be considered a physical science, is in this respect analogous in its methods of reasoning to the moral sciences; the observed phenomena serve to indicate the great Mind of the universe, and to expound its acts, as the circumstances of a human agent serve to interpret his conduct.

(66) 'Die Aufgabe des Geschichtschreibers ist die Darstellung des Geschehenen. Das Geschehene aber ist nur zum Theil in der Sinnenwelt sichtbar, das Uebrige muss hinzu empfunden, geschlossen, errathen werden. Was davon erscheint, ist zerstreut, abgerissen, vereinzelt; was dies Stückwerk verbindet, das Einzelne in sein wahres Licht stellt, dem Ganzen Gestalt giebt, bleibt der unmittelbaren Beobachtung entrückt. Sie kann nur die einander begleitenden und auf einander folgenden Umstände wahrnehmen, nicht den innern ursachlichen Zusammenhang selbst, auf dem doch allein auch die innere Wahrheit beruht.'-Wilhelm v. Humboldt, Ueber die Aufgabe des Geschichtschreibers, Gesammelte Werke, vol. i. p. 1.

152.

CHAPTER VI.

ON THE APPLICABILITY OF THE METHOD OF

IT

EXPERIMENT TO POLITICS.

has been shown in the preceding chapter, what are the facts upon which the science and practice of Politics are founded, and what are the means employed for their observation. We have learned that, however numerous, complex, and variable these facts may be, they lie within the range of simple observation, and admit of being determined with certainty. It is true that fraud, or concealment, or neglect, or inaccuracy of observation, may give rise to doubt and controversy about particular events, and if the extant evidence to a historical fact is defective, that defect can never be supplied: but all political facts lie within the sphere of our senses, and if due attention is paid to them at the time of their occurrence, they always admit of being observed. For so much of a political fact as consists in mental operations, which are necessarily concealed from our view, we can only, as in ethics, proceed by way of inference from outward and visible marks. This, indeed, is a source of obscurity not peculiar to politics, but common to all departments of knowledge which involve human action.

It is, however, in politics, as in physics: a part, and only a small part, of the process of investigation consists in the mere collection and registration of facts. All intelligent observation must be made with relation to some end; the observer has some ground for his preference, some criterion for discriminating between the facts to be noted, and the facts to be disregarded.(') In general, he seeks to ascertain the connexion of certain facts, in

(1) That all observation ought to be directed by a theory is shown by M. Comte, Cours de Phil. Pos. tom. iv. p. 418-25, 665-6.

the way of cause and effect. Now, in attempting to solve a problem of political causation, we are in general presented with a large number of contemporary events, some of which are selected, according to the peculiar views of individuals, as the antecedents of a given phenomenon. If, for example, agricultural, manufacturing, or commercial distress prevails in any country, there will be different modes of accounting for it, according to the comparative intelligence and knowledge, the economical interests, or the political opinions and connexions of different persons. Some will trace it to one source, and some

to another; much controversy and dissension will arise, and the question will only be set at rest by the succession of new and more pressing disputes on similar subjects.

Now if, for the purpose of determining doubtful questions of this sort, politics possessed some summary and decisive method, some means of searching out the concatenation of phenomena, and of discriminating between true and false causes, similar to that which several of the physical sciences derive from the use of EXPERIMENTATION, (2) a powerful instrument of reasoning would be placed in the hands of the political inquirer.

Let us, therefore, consider what is the real use of experiments in the physical sciences, and how far this method of investigating truth is applicable to the sciences, of which man, and not matter, is the subject.

§ 2 Experiments in physics are of two sorts, the experiments of science and the experiments of art.

The former have

The former are

a philosophical; the latter, a practical purpose. intended to establish truth, or to refute error: the latter are intended to ascertain, by actual trial, the working qualities of some tool, instrument, machine, or mechanical contrivance; the influence of some remedy on the human frame; the wholesome

(2) On the subject of experiment in the physical sciences, see Playfair's Dissertation in the Enc. Brit. vol. i. p. 466, ed. 7: Herschel's Discourse on the Study of Natural Philosophy, § 67: Cuvier, Règne Animal, tom. i. p. 4-6; Comte, Cours de Philosophie Positive, tom. ii. p. 404-7; tom. iii. p. 28, 320, 333; Mill's Logic, vol. i. p. 441-4.

ness of some food for man or animal; the effects of some mode of cultivation upon plants; the strength of materials; the comparative speed of animals or ships; and the like.

Examples of the former class are the experiments of mechanics, chemistry, optics, and of the sciences of electricity, magnetism, and heat. The experiments of Archimedes on the lever, of Galileo on falling bodies, of Newton on light, of Franklin on the lightning, and of Lavoisier on combustion, are celebrated instances of this class. (3) The latter class of experiments are the experiments of the useful arts; all those trials which inventors make when they seek to apply the principle of their invention in practice; when they confront their idea with realities, and seek to convert their speculative contrivance into a working machine. That many such ideas are formed which meet with no practical success, that the inventor often deludes himself with impracticable plans, or is wanting in mechanical skill to give effect to his own principle, is proved by the large number of patents for new inventions which lie barren in the hands of the patentee. Of the same nature are the experiments made by the agriculturist, the breeder of cattle, the planter, the gardener, for the sake of discovering what are the circumstances under which the objects of his care will thrive most, and yield the largest profit.

The former of these classes of experiments appears to correspond with the experimenta lucifera-the latter, with the experimenta fructifera, of Bacon.(*) Scientific experiments imply a theoretical conclusion. They always prove something. No man makes a philosophical experiment (unless it be merely for the negative purpose of confutation), without inferring from it a law

(3) See Whewell, Hist. of Ind. Sciences, b. ix. c. 3; b. xi. c. 1.

(4) See Nov. Org. præfat. p. 156. lib. i. aph. 99, 121, lib. ii. aph. 36. ad fin: also, the Parasceue ad Historiam Naturalem et Experimen talem, vol. xi. p. 408. Artium experimenta' are mentioned in Nov. Org.

i. 103.

In the treatise De Augmentis, v. 2, invention or discovery is divided into two branches-experientia litterata, and interpretatio naturæ (vol. viii. p. 264. 276). These nearly correspond to practical and scientific experiments, although the experientia litterata is not confined to the former class. Compare Adv. of Learning, vol. ii. p. 182.