We have found out that the distortion in dreams which hinders our understanding of them is due to the activities of a censorship, directed against the unacceptable, unconscious wish-impulses. But of course we have not asserted that the censorship is the only factor responsible for the distortion, and as a matter of fact a further study of dreams leads to the discovery that there are yet other causes contributing to this effect; that is as much as to say, if the censorship were eliminated we should nevertheless be unable to understand dreams, nor would the manifest dream be identical with the latent dream-thoughts. This other cause of the obscurity of dreams, this additional contribution to distortion, is revealed by our becoming aware of a gap in our technique. I have already admitted to you that there are occasions when persons being analysed really have no associations to single elements in their dreams. To be sure, this does not happen as often as they declare that it does; in very many instances the association may yet be elicited by perseverance; but still there remain a certain number of cases where association fails altogether or, if something is finally extorted, it is not what we need. If this happens during psycho-analytic treatment it has a certain significance which does not concern us here; but it also occurs in the course of interpretation of dreams in normal people, or when we are interpreting our own. When we are convinced in such circumstances that no amount of pressing is of any use, we finally discover that this unwelcome contingency regularly presents itself where special dream-elements are in question; and we begin to recognize the operation of some new principle, whereas at first we thought we had only come across an exceptional case in which our technique had failed. In this way it comes about that we try to interpret these “silent” elements, and attempt to translate them by drawing upon our own resources. It cannot fail to strike us that we arrive at a satisfactory meaning in every instance in which we venture on this substitution, whereas the dream remains meaningless and disconnected as long as we do not resolve to use this method. The accumulation of many exactly similar instances then affords us the required certainty, our experiment having been tried at first with considerable diffidence. I am presenting all this somewhat in outline, but that is surely allowable for purposes of instruction, nor is it falsified by so doing, but merely made simpler. We arrive in this way at constant translations for a series of dream-elements, just as in popular books on dreams we find such translations for everything that occurs in dreams. You will not have forgotten that when we employ the method of free association such constant substitutions for dream-elements never make their appearance.