Freud has pointed out in many places[150] with unmistakable clearness that the psychosexual relationship of the child towards his parents, particularly towards the father, possesses an overwhelming importance in the content of any later neurosis. This relationship is in fact the infantile channel _par excellence_ in which the libido flows back[151] when it encounters any obstacles in later years, thus revivifying long-forgotten dreams of childhood. It is ever so in life when we draw back before too great an obstacle--the menace of some severe disappointment or the risk of some too far-reaching decision--the energy stored up for the solution of the task flows back impotent; the by-streams once relinquished as inadequate are again filled up. He who has missed the happiness of woman's love falls back, as a substitute, upon some gushing friendship, upon masturbation, upon religiosity; should he be a neurotic he plunges still further back into the conditions of childhood which have never been quite forsaken, to which even the normal is fettered by more than one link--he returns to the relationship to father and mother. Every psychoanalysis carried out at all thoroughly shows this regression more or less plainly. One peculiarity which stands out in the works and views of Freud is that the relationship to the father is seen to possess an overwhelming importance. This importance of the father in the moulding of the child's psycho-sexuality may also be discovered in a quite other and remote field, in the investigation of the family.[152] The most recent thorough investigations demonstrate the predominating influence of the father often lasting for centuries. The mother seems of less importance in the family.[153] If this is true for heredity on the physical side how much more should we expect from the psychological influences emanating from the father? These experiences, and those gained more particularly in an analysis carried out conjointly with Dr. Otto Gross, have impressed upon me the soundness of this view. The problem has been considerably advanced and deepened by the investigations of my pupil, Dr. Emma Fürst, into familial resemblances in the reaction-type.[154] Fürst made association experiments on one hundred persons belonging to twenty-four families. Of this extensive material, only the results in nine families and thirty-seven persons (all uneducated) have been worked out and published. But the painstaking calculations do already permit some valuable conclusions. The associations are classified on the KRÆPELIN-ASCHAFFENBURG scheme as simplified and modified by myself; the difference is then calculated between each group of qualities of the subjects experimented upon and the corresponding group of every other subject experimented upon. Thus we finally get the differentiation of the mean in reaction-type. The following is the result:--