past the great cities, illustrates how clearly everybody sees natural law at work in society. It is the laws of society that determine the direction in which population moves. For example, in peoples at all advanced the head of navigation of rivers is usually the site of the principal towns. A short time ago, when water was more used than now as a power, there was usually combined with the advantages offered by the head of navigation (vessels being then smaller than now), the additional advantage of the fall in the stream, which is almost always greatest at the point where the piedmont plateau joins the coastal plain. As streams only reach their base level after emerging upon the coastal plain, this sudden fall almost always occurs a short distance above the head of navigation. As this is true of all the streams that drain a continent, a line may be drawn through this point on all the rivers and it will be approximately parallel to the coast. Such a line is called the 'fall line,' and it is a law of population that the first settlements of any country take place along the fall line of its rivers.
There are many laws that can be similarly illustrated, and careful observation reveals the fact that all social phenomena are the results of its laws. The one example given must suffice in the present case. But these social, or sociological, laws may themselves be grouped and generalized, and higher laws discovered. If we carry the process far enough we arrive at last at the fundamental law of everything psychic, especially of everything affected by intelligence. This is the law of parsimony. It has, as we shall see, its applications in biology, and its homologue in cosmology, but it was first clearly grasped by the political economists, and by many it is regarded as only an economic law. Here it is usually called the law of greatest gain for least effort, and is the basis of scientific economics. But it is much broader than this, and not only plays an important rôle in psychology, but becomes, in that collective psychology which constitutes so nearly the whole of sociology, the scientific corner stone of that science also.
We have seen that the quality of scientific exactness in sociology can only be clearly perceived in some of its higher generalizations, where, neglecting the smaller unities which make its phenomena so exceedingly complex, and dealing only with the large composite unities that the minor ones combine to create, we are able to handle the subject, as it were, in bulk. Here we can plainly see the relations and can be sure of their absolute uniformity and reliability. When we reach the law of parsimony we seem to have attained the maximum stage of generalization, and here we have a law as exact as any in physics or astronomy. It is, for example, perfectly safe to assume that under any and all conceivable circumstances, a sentient and rational being will always seek the greatest gain, or the maximum resultant of gain—his 'marginal' advantage.
