But the original investigator assigns reasons, and with care the children may be allowed to attempt that. This, however, should not be carried far; incorrect explanations should be criticised; and the class should at length be given all the elements of the correct explanation which they have not determined for themselves. Later, pupils should be encouraged to name related phenomena, to mention things which they have seen happen which are due to associated causes, and to suggest variations for the experiment and tests for its explanation. Good results may be made to follow this kind of work even with very young pupils. A child grows in mental strength by using the powers he has, and mistakes seen to be such are not only steps toward a correct view of the subject under consideration, but are steps toward that habit of mind which spontaneously presents correct views at once in study which comes later in life. Another method is this: The pupil may know what is expected to happen, as well as the conditions given, and held responsible for an observation of what does happen and a comparison of what he really observes with what he expects to observe. Explanations are usually given a class, often in books with which they are furnished, instead of being drawn from them, in whole or in part, by questioning, when physical science is studied in this way. Indeed, this method is a necessity when text books are used, unless experiments from some outside source are introduced. Who shall perform the experiments? With young pupils everywhere, and in most of our common, and even in many of our graded schools, the experiments must be performed by the teacher. With young pupils the time is too limited, and the responsibility and necessary care too great to permit of any other plan being practical. In many of our schools the small supply of apparatus renders this necessary even with larger pupils. Added to the reasons already given is the important one that in no other way--by no other plan--can the teacher be as readily sure that his pupils observe and reason fully for themselves. In this normal school a course in physics, in which the experiments are all performed in the class room by the teacher, is followed by a course in chemistry, in which the members of the class perform the experiments for themselves in the laboratory. And, notwithstanding the age, maturity, and previous observation of the pupils, a great deal must be done both in the laboratory and in the recitation room to be sure that all that happens is seen--that the purpose is clearly held in the mind--that the reason is fully understood. With older pupils and greater facilities, however, the experiments should be performed by the pupils themselves. Constant watchfulness is necessary, it is true, to insure to the pupil the full educational value of the experiment. With this watchfulness it can be done, and the advantages are numerous. Among them are: