The last of the three methods is by far the best, whether the memory desired be rote or logical, for several reasons.  In the first place it involves both the other methods or goes beyond them.  Second, it is economical, for the learner knows when he knows the lesson.  Third, it is sure, for it establishes connections as they will be used—­in other words, the learning provides for recall, which is the thing desired, whereas the other two methods establish only connections of impression.  Fourth, it tends to establish habits that are of themselves worth while, such as assuming responsibility for getting results, testing one’s own power and others.  Fifth, it encourages the use of the two factors upon which memory depends, which are most capable of development, i.e., number and organization of associations.

In connection with the use of the material two methods have been employed—­the part method and the whole method.  The learner may break the material up into sections, and study just one, then the next, and so on, or he may take all the material and go through with it from the beginning to the end and then back again.  Experimental results show the whole method to be the better of the two.  However, in actual practice, especially with school children, probably a combination of the two is still better, because of certain difficulties arising from the exclusive use of the whole method.  The advantages of the whole method are that it forms the right connections and emphasizes the complete thought and therefore saves time and gives the right perspective.  Its difficulties are that the material is not all of equal difficulty and therefore it is wasteful to put the same amount of time on all parts; it is discouraging to the learner, as no part may be raised above the threshold of recall at the first study period (particularly true if it is rote memory); it is difficult to use recall, if the whole method is rigidly adhered to.  A combination of the two is therefore wise.  The learner should be encouraged to go over the material from beginning to end, until the difficult parts become apparent, then to concentrate on these parts for a time and again go over from the beginning—­using recall whenever possible.

A consideration of the time element involved in memorizing has given use to two other methods, the so-called concentrated and distributive.  Given a certain amount of time to spend on a certain subject, the learner may distribute it in almost an infinite number of ways, varying not only the length of the period of practice, but also the length of time elapsing between periods.  The experimental work done in connection with these methods has not resulted in agreement.  No doubt there is an optimum length of period for practice and an optimum interval, but too many factors enter in to make any one statement.  “The experimental results justify in a rough way the avoidance of very long practice periods and of very short intervals.