falling bodies does not describe the way in which bodies actually behave with reference to the earth. Bodies never did and never will behave in this way under natural or uncontrolled conditions. Hence, the law of falling bodies is not a "natural" law in the sense of being man's conscious descriptive reproduction of a constant or fixed relationship in nature. It is a human, a man-made law or norm, constructed out of a multitude of specific experiences to aid man in determining the behavior of any particular object with reference to the earth and other objects, but not of all objects uniformly and absolutely or invariably. Without such a law there could be no accurate measurement of the path of projectiles or of many other terrestrial physical events or physical mass movements. But its utility consists in its establishment of a point of departure or norm from which to calculate variations from this norm for each particular gravitational event under the circumstances of atmospheric movement or other modifying conditioning pressure and which obtains, not for the measurement of all events in uniform conformity to or identity with the law.

When the physicist says that this law tells us how masses would behave "naturally" and that it is therefore nature's law, though subject to interruption of uniformity in practice, he is speaking, not as a scientist, but as a metaphysician. It is not as if there is but one law of falling bodies. There may be an infinite number of such laws, the actual number of the laws depending on the conditions for the generalization of which they were formulated. The one we use in physics was made for highly artificial conditions, which never occur in nature, except under the experimental control of man. It describes the way in which bodies fall in a vacuum. All other laws of falling bodies will have different formulas or incremential ratios according to the degree to which the medium through which they fall varies from an absolute vacuum, and also to the extent to which this medium is in motion or not at rest. The physicists arbitrarily fixed upon the condition of an absolute vacuum as a convenient basis for generalization and formulation, and they were probably wise in doing so for the sake of securing effective results in applying the norm to concrete actual conditions. It gave them the simplest formula for a law of falling

bodies of which we can conceive. In removing the differential factors of atmospheric and other pressures through a reduction of the conditions of operation to those of a vacuum the experimenter hit empirically upon one of the best-known practices of statistical correlation since discovered and formulated as a principle.

Other laws of falling bodies would also have served as norms for computing variations in practice and thus have enabled the mathematical physicist to control the part of his world which could be viewed or comprehended through a law of falling bodies, but the computations would have been more difficult and therefore the control of adjustment to his world would have been less accurate and economical. It is apparent therefore that the most economical law of falling bodies, because the one from which most variables were eliminated in its formulation, was selected by the physicists for the purpose of this control. We may find a somewhat analogous case in our arithmetic. We use the decimal system for the purpose of representing abstract multiple concepts of numbers, more complex than can be grasped by direct concrete perceptions. Some mathematicians maintain that the duo-decimal system would have been more economical for this purpose of aggregating multiple numerical values, because the mathematical operations would have been more rapidly and effectively performed in higher computations. But the human hands had only ten digits, consequently we learned to use the decimal instead of the duo-decimal system, and now custom and the mass of ready-made computations prevents our easily changing the system. What we might do better with the one system we do tolerably with the other. Similarly, what we do with one law of falling bodies we might do better or less well with another. The point is, however, that there is not just one law of falling bodies possible, nor has the one we use the right to claim to be a "natural" law on the assumption that it is an accurate picture or description of what actually occurs in nature. Each and every one of these possible laws of falling bodies is or would be man's way of seeing certain aspects of nature abstractly, of looking perceived phenomena into mathematical perspective according to his point of view or position in the world of physical phenomena -nothing less and nothing more.

Laws are usually formulated initially and in their rudimentary forms as the result of a concrete problem or set of problems in a practical adjustment situation. Consequently, the form which a scientific law takes in the process of formulation is likely to be the one most economical for man in controlling his world or in making adjustments to it. Sometimes the variant and less economical forms of laws would be so far removed from the conditions under which they could be utilized by man that they would not be practicable, although they would be true for the conditions with reference to which they were formulated or for any other conditions by making suitable allowance for variations of actual condition from those assumed by the projected and simplified norm or law. Thus we ordinarily state the formula for the production of an acid on the assumption that the compounds are chemically pure, but it might be stated on the assumption of any degree of chemical impurity for any or all contributing compounds. However, it is more economical in practice to establish a norm at the base of hypothetical chemical purity and certain fixed conditions of temperature, atmospheric pressure, etc., and allow for variations, than it would be to state the formula for particular grades of impurity of chemicals and variations of pressure and temperature, which would seldom if ever be repeated accurately, and then compute the variation of the actual state of affairs from the basic norm or law of metathesis.

The formulation of social laws is subject to the same general principles as is the formulation of physical and chemical laws. There is no difference in the form of a social and a physical or a chemical law. But there is usually a great difference in the degrees of accuracy of formulation of laws in the two fields. The earliest generalizations in both fields are vague and indefinite, not quantitatively stated or mathematically determined. Physics and the exact sciences generally achieved the stage of stating laws in quantitative formulas considerably before the social sciences reached the same stage. The reason for this is clear. Physical phenomena are much more simple, that is, directly apprehensible in form and occurrence by the senses, and are therefore much more easily looked into relationship or perspective and into quantitative abstract integration by the scientist. Physical phenomena are vari

able primarily in only one dimension, with relation to one another, as masses; that is, there is little, at least very slow, modification within the masses. But social phenomena vary rapidly and largely in two dimensions, both with respect to one another as persons and within the organization of the personalities or institutions and processes themselves. People have minds and physiological processes, and groups are continually undergoing reorganization, and both people and groups must adjust to others of the same or different kinds. The technique of the adjustment is all the more difficult because of this internal instability. Hence, the greater difficulty experienced by the scientist in seeing social phenomena and processes as wholes and of formulating these processes abstractly in the form of definite laws or quantitative generalizations.

But the problem is not an impossible one, as is proved by the success of the statisticians of death expectancies and the construction of life-tables for the use of actuaries. The economists, among the social scientists, in the fields of finance and markets, have made most progress in arriving at quantitative generalizations, probably because they have been able to handle such a vast mass of cases that they could apply successfully their method of generalizing statistically or mathematically. This method of quantitative generalization employed by the social scientist must be by means of the forms of mathematics commonly called multiple and partial correlation and the theory of probabilities. Simpler mathematical processes would not be adequate to the masses of data which are necessary to eliminate the exceptional instances. It is reasonable to anticipate that ultimately all social phenomena which occur in vast numbers may be viewed into some unitary system as formulas, partaking of the nature of social laws, just as the phenomena of physics are capable of being so viewed or interpreted. But not only is the problem of generalizing quantitatively more difficult for the social sciences, but also the problem of the application of the social laws, that is, the problem of computing adjustments of social phenomena on the basis of variations in the concrete instances from the norm or law, must also be more difficult than in the case of the variation of concrete physical phenomena from the physical norm or law.

If there is this greater difficulty in the formulation of social laws, because of the greater complexity and variability of social phenomena, how much greater would be the task of formulating a general law of social progress. As Mr. Shafer remarked, it is not possible to formulate such a general law of progress—at least not for the present. This, however, is quite another thing from maintaining that there is or can be no such thing as social progress. It only means that we cannot be certain whether any particular policy or event makes for progress, because we have no universal and quantitatively established norm with which to measure its value. However, if we have no accurate universal law of social progress, we may still find some guideposts on the way. Because man cannot yet see the whole meaning of his existence is no adequate reason for supposing that he may not be able to perceive some parts of it and evaluate policies with varying degrees of accuracy and effectiveness. Knowledge is cumulative and from it we gradually get perspective; that is, that generalization of viewpoint which is the essence and basis of law. The denial of progress itself, in regard to the whole or any part of the behavior of man, in which the critics of the theory of social progress so fully indulge, is in itself a tacit acknowledgment of this fact of the existence of social value norms, at least on a limited scale, and constitutes an attempt to utilize them in defending a thesis or point of view. That is, it is itself an attempt to look facts or phenomena into a generalized statement or to establish a norm as a basis for the measurement or evaluation of particular events.

The biologists have not yet been able to formulate a general law of evolution with anything like mathematical exactness, either for the whole process of evolution or for any special aspect of it, such as organic evolution. Yet a special phase of evolution is presumably less complex than the whole range of social progress, which would necessarily involve the adjustment of all classes of social phenomena. All accounts of evolution are as yet verbally descriptive and even these descriptions probably have not yet reached their final statement. The scientist recognizes that he must begin, in the formulation of laws, with phenomena in more circumscribed and immediate fields. Only gradually can he hope to find a firm