In observation we simply _watch_ the phenomenon; in experiment we _make_ it. In experiment we not only observe, but we _manipulate_ the circumstances so as to present the phenomenon under the most favorable conditions for observation. “In observation,” says Mill, “we _find_ an instance in nature suited to our purposes”; whilst in experiment, by an artificial arrangement of circumstances, we _make_ an instance suited to our purpose. In observation we watch for causes; in experiment we work for effects. We may thus define experiment as _the act of making phenomena occur for the purpose of watching for effects_. In experiment there is much which is merely observation. In fact experiment _is_ observation in which the phenomenon is artificially produced. For the sake of definiteness, however, any observation which involves a _manipulation of circumstances_, may be designated as _experimental_. =4. RULES FOR LOGICAL OBSERVATION AND EXPERIMENT.= To the uninitiated, the matter of observation seems an easy task, and yet when one hears two honest men swear to diametrically opposite facts which have come to them from observing the same phenomenon, his faith is shaken. “Eyes have they but see not” is a logical truth as well as a moral one. Only the observation of the trained can be depended upon; and yet this should not discourage the layman, for even he, by a little conscientious effort towards careful observation, may greatly increase his store of accurate knowledge and add to the joy of living. The attending rules are usually heeded by the trained scientist in matters of observation and experiment: _First Rule. The observations should be precise._ The time, the place, the surrounding conditions _must_ be accurately noted. Many artificial contrivances have been devised because of the desire of the scientist to be precise. Instruments like the balance, the thermometer, the microscope, etc., has he invented, and various devices and methods has he adopted for the sake of _precision_. A common method is to take an _average of observations_. For example, to estimate justly the class work of a student, the teacher should not be content with the ratings of one or two recitations, but must average the ratings of many recitations. Again, a child may be led to discover approximately the value of the sum of the interior angles of a triangle by measuring the angles of many triangles and then striking an average; assume that the following results are obtained by such a procedure: (1) 178, (2) 181, (3) 179, (4) 180, (5) 182; adding these and dividing by 5 gives 180.