I pass over the striking circumstance that here, as elsewhere in the analysis of dreams, associations of the most widely different values are employed for the establishment of thought connections as though they were equivalent, and I yield to the temptation to regard the process by which _amyls_ in the dream thoughts are replaced by _propyls_, as though it were plastic in the dream content. On the one hand is the chain of ideas about my friend Otto, who does not understand me, who thinks I am in the wrong, and who gives me the cordial that smells like amyls; on the other the chain of ideas—connected with the first by contrast—about my friend William, who understands me and who would always think I was in the right, and to whom I am indebted for so much valuable information about the chemistry of the sexual processes. Those characteristics of the associations centering about Otto which ought particularly to attract my attention are determined by the recent occasions which are responsible for the dream; _amyls_ belong to these elements so determined which are destined to get into the dream content. The group of associations “William” is distinctly vivified by the contrast to Otto, and the elements in it which correspond to those already excited in the “Otto” associations are thrown into relief. In this whole dream I am continually referring to a person who excites my displeasure and to another person whom I can oppose to him or her at will, and I conjure up the friend as against the enemy, feature for feature. Thus amyls in the Otto-group suggests recollections in the other group belonging to chemistry; trimethylamin, which receives support from several quarters, finds its way into the dream content. “Amyls,” too, might have got into the dream content without undergoing change, but it yields to the influence of the “William” group of associations, owing to the fact that an element which is capable of furnishing a double determination for amyls is sought out from the whole range of recollections which the name “William” covers. The association “propyls” lies in the neighbourhood of _amyls_; Munich with the propylæum comes to meet _amyls_ from the series of associations belonging to “William.” Both groups are united in _propyls—propylœum_. As though by a compromise, this intermediary element gets into the dream content. Here a _common mean_ which permits of a manifold determination has been created. It thus becomes perfectly obvious that manifold determination must facilitate penetration into the dream content. A displacement of attention from what is really intended to something lying near in the associations has thoughtlessly taken place, for the sake of this mean-formation.