THE STUDY OF MENTAL CONDITIONS IS A MOST POWERFUL ALLY OF ETHICS. THE HELL TO BE ENDURED HEREAFTER, OF WHICH THEOLOGY TELLS, IS NO WORSE THAN THE HELL WE MAKE FOR OURSELVES IN THIS WORLD BY HABITUALLY FASHIONING OUR CHARACTERS IN THE WRONG WAY. COULD THE YOUNG BUT REALIZE HOW SOON THEY WILL BECOME MERE WALKING BUNDLES OF HABITS, THEY WOULD GIVE MORE HEED TO THEIR CONDUCT WHILE IN THE PLASTIC STATE. EVERY SMALLEST STROKE OF VIRTUE OR OF VICE LEAVES ITS NEVER SO LITTLE SCAR. THE DRUNKEN RIP VAN WINKLE, IN JEFFERSON'S PLAY, EXCUSES HIMSELF FOR EVERY FRESH DERELICTION BY SAYING, ‘I WON'T COUNT THIS TIME!' WELL! HE MAY NOT COUNT IT, AND A KIND HEAVEN MAY NOT COUNT IT; BUT IT IS BEING COUNTED NONE THE LESS. DOWN AMONG HIS NERVE CELLS AND FIBERS THE MOLECULES ARE COUNTING IT, REGISTERING AND STORING IT UP TO BE USED AGAINST HIM WHEN THE NEXT TEMPTATION COMES. NOTHING WE EVER DO IS, IN STRICT SCIENTIFIC LITERALNESS, WIPED OUT. OF COURSE, THIS HAS ITS GOOD SIDE AS WELL AS ITS BAD ONE. AS WE BECOME PERMANENT DRUNKARDS BY SO MANY SEPARATE DRINKS, SO WE BECOME SAINTS IN THE MORAL, AND AUTHORITIES AND EXPERTS IN THE PRACTICAL AND SCIENTIFIC SPHERES, BY SO MANY SEPARATE ACTS AND HOURS OF WORK. Table XII shows the method which was used in deriving 10 backward lists and 10 separated lists from the prose selection. The first 3 and the last 13 words of the prose passage are omitted in the table, which contains 200 words arranged in the same order as in the prose selection. The first backward list consists of the 10 words in the row to the left of "B" in the table, arranged and written in the reverse order, as "BECOME WE SO DRINKS SEPARATE MANY SO BY DRUNKARDS PERMANENT." Each of the remaining B-lists, 2B, 3B, etc., is constructed in the same way. The first separated list is made up of the 10 italicized words in the column above "IS" in the table, reading from bottom to top, as "THE BECOME STRICT AND IT I NEVER MORE WRONG IS," etc. PROCEDURE The procedure which we have devised may be divided into 3 parts: (1) Learning the 20 B- and S-lists, (2) Learning the prose passage, and (3) Relearning the 20 B- and S-lists. It is possible in this way to compare the efficiency of learning the B- and S-lists, both before and after learning the prose passage. Learning the B- and S-lists.-The writer acted as experimenter, and the subjects were taken individually. The subject first learned three preliminary practice lists, thereby becoming acquainted with the procedure. The method of learning was similar to that used with the test lists in Part I, except that the efficiency of learning was measured in all cases in terms of the time required to learn each list to the point of one correct and ready recitation. The subject was then required to learn the 20 B- and Slists singly, in the order 1B, 1S, 2B, 2S, etc. An effort was made to control the conditions in such a way that the subject should not know that the B-lists were made up of words taken from a logical passage and arranged in the reverse order. At the end of the one-hour period required for this procedure, the subject was questioned about the construction of the lists their relative difficulty, etc. No subject noticed the method of construction of any of the lists. The time measures obtained make it possible to compare the natural difficulty which each subject experienced in learning the Band S-lists before learning the prose passage. Learning the prose passage.-Each of the 31 subjects was taken for the one-hour period in the individual procedure described above between the Monday and Saturday of one week. When the class met on the following Monday, printed copies of the prose passage were distributed, and the group of subjects was instructed to take the passage home and learn it. On the Tuesday, Wednesday, Thursday, and Friday of this week, the first 10-minutes of the class-period were used in having the subjects recite the prose passage to each other in pairs. At the end of the second week, the subjects were urged to continue practicing the prose passage every day, so that they could recite it without error during the third and last week of the experiment. It is well to point out that logical material has certain very definite advantages over rote material for the purposes of the present experiment. When the rote material (of Part I) is used, the subject's thought processes wander and fluctuate on account of the general lack of interest and meaning in the disconnected items. It is probably not possible to control the thinking processes in practice with such material. But the case is somewhat different with the logical material. After the prose passage has been learned to the point of one correct recitation, further repetition results in a condition in which the process of overlearning is admirably controlled. The subject recognizes his own errors, and promptly makes the desired corrections himself. He is held in line in a form of practice which is more or less automatically controlled. As the controlled practice continues, the overlearning becomes more and more pronounced, and, as we shall see later, the weak remote associations formed by the irregularities in the early stages of the practice are practically forgotten. Relearning the B- and S-lists.-Each subject was scheduled for a one-hour period exactly 2 weeks after the first learning of the B- and S-lists. He was first asked to recite the prose passage. No comments were made on his recitation, and his errors were not corrected. The subject then relearned the B- and S-lists, using the same method as before. When the third and last part of the procedure had been completed, the subject was asked a number of questions which were designed to find out whether he had any insight into the problem being investigated. He was asked whether he noticed the method of construction of the 20 lists, how many of the words had been taken from the prose passage, whether the lists could be divided into several classes according to any criterion, why some lists were harder to learn than others, the reason for the soft spots and the difficult places, etc. Special care was taken not to suggest the method of constructing the 20 lists and the general purpose of the experiment. Most of the subjects thought that only about three-fourths of the words in the 20 lists had been taken from the prose passage. No subject suspected the method of constructing the B- and S-lists from the prose passage. The writer also asked the subjects for their opinion in regard to the general purpose of the experiment, and the reason for using this type of material. Perhaps the most significant feature of our procedure is that no subject had any insight into the problem being investigated. RESULTS Influence of learning prose passage on backward association: individual results.-26 out of a total of 31 members of the class in elementary psychology completed the procedure as we have described it above. The individual results for these 26 subjects are given in Tables XIIIA-D. The general method of treating the data may be illustrated by reference to the results of the first subject, shown in Table XIIIA. The first column in the Table (A) shows that this subject required an average of about 39 sec to learn the 10 B-lists, and an average of about 41 sec to learn the 10 S-lists. The superiority of the B-lists over the S-lists amounts to 2 sec. This difference is expressed in per cent. by dividing 2 by the average of 39 and 41. After the learning of the prose passage, the same 10 B-lists required an average learning time of 17 sec, and the same 10 S-lists required an average learning time of 28 sec. difference is now + II sec, in favor of the B-lists. The The superiority of the B-lists over the S-lists is greater after learning the prose passage than before. The difference (d-c) — (b-a) is + 9 sec, or + 43 per cent. It appears from the results of this subject that learning the prose passage has resulted in the formation of associations in the backward direction. The difference after learning (d-c) is more significant than the difference before learning (b-a) on account of the shift in the base of comparison. For most purposes, the difference (d-c) - (b-a) is not as important as the percentage difference given in the lower row of the table. If the figures in the lower row are positive, the results are in favor of backward association; if these figures are negative the results indicate that the formation of associations in the backward direction has been inhibited. Tables XIIIA-D show a variety of facts in regard to the individual tendencies of the 26 subjects. The value for b-a is positive in 18 cases out of 26; and the value for d-c is positive in 14 cases of out 26. This indicates that the effect of learning the prose passage is to inhibit the formation of |