Page:Popular Science Monthly Volume 86.djvu/172

168 extensively used text-book of zoology, written by a biologist of international reputation, occurs this passage:

An excised representative sample of hydra will reproduce the whole: but you can not perform this experiment with the frog.

To one who looks upon an experiment as a means of testing hypotheses there is no obvious reason why "this experiment" can not be performed with a frog, or any other beast. But if the experiment is a means for getting certain desired objective results, of course it is impossible to get a complete frog to regenerate from an "excised representative sample"—as we know from experiments!

That the experiment does not always mean to the teacher the same as it does to the investigator may be inferred from the fact that many teachers are not averse to "faking" experiments that are arranged for demonstration purposes. William James tells in one of his papers of his own performance in a physiological demonstration, and he justifies it upon pragmatic grounds. The question I am raising is not one of ethics, but of clear thinking. If the experiment is a didactic tool for presenting concrete, objective processes, it falls into the same category as wall-charts and models. The demonstration experiment need not then be any more "real" than a glass model of the eye or of a diamond. But if the experiment is used by the teacher for the purpose of teaching method in thought or in the solution of problems, the "unsuccessful" experiment should be at least as illuminating and educating as the "successful" one.

A third source of confusion lies in the apparently harmless little word "law." A student of science should certainly know what is meant by a "law of nature"—but we may not expect him to if his teacher does not. Now it is altogether too common to hear teachers of science speak of the laws of nature in exactly the same way as ordinary folks whose notion of "law" is derived from the statutes of the commonwealth or the commandments of the gods. In science a law is presumably a generalization from a limited series of experiential data, not a prohibitive or mandatory order from some superior authority. Our attitude toward Boyle's law, for example, is in no way related to our loyalty to Mr. Boyle. In ordinary usage there may be violations of "law" and such violations are frequently followed by disagreeable consequences. But in "nature" the consequence is not something superimposed by way of punishment or retribution; it is itself a part of the law, and integral in the general process formulated in the law. Laws of nature can not be violated in the sense that statutory laws can be. Laws of health are descriptive generalizations of the conditions under which normal health is maintained. Yet we speak of empirical rules for securing these conditions as also being "laws of health." In practise we may or we may not observe these rules; but we can not violate the laws. Morbid conditions also arise in conformity to law. There is