Displacement of accent is not a legitimate device if our object be the expression of thought; though we do sometimes admit it in waking life in order to produce a comic effect. I can to some extent convey to you the impression of confusion which then results, by reminding you of an anecdote, according to which there was in a certain village a smith who had committed a capital offence. The court decided that the smith was guilty; but, since he was the only one of his trade in the village and therefore indispensable, whereas there were three tailors living there, one of these three was hanged in his place! The third achievement of the dream-work is the most interesting from the psychological point of view. It consists in the transformation of thoughts into _visual images_. Let us be quite clear that not everything in the dream-thoughts is thus transformed; much keeps its original form and appears also in the manifest dream as thought or knowledge, on the part of the dreamer; again, translation of them into visual images is not the only possible transformation of thoughts. But it is nevertheless the essential feature in the formation of dreams, and, as we know, this part of the dream-work is, if we except one other case, the least subject to variation; for single dream-elements, moreover, _plastic word-representation_ is a process already familiar to us. Obviously this achievement is by no means an easy one. In order to get some idea of its difficulty, imagine that you had undertaken to replace a political leading article in a newspaper by a series of illustrations; you would have to abandon alphabetic characters in favour of hieroglyphics. The people and concrete objects mentioned in the article could be easily represented, perhaps even more satisfactorily, in pictorial form; but you would expect to meet with difficulties when you came to the portrayal of all the abstract words and all those parts of speech which indicate relations between the various thoughts, e.g. particles, conjunctions, and so forth. With the abstract words you would employ all manner of devices: for instance, you would try to render the text of the article into other words, more unfamiliar perhaps, but made up of parts more concrete and therefore more capable of such representation. This will remind you of the fact that most abstract words were originally concrete, their original significance having faded; and therefore you will fall back on the original concrete meaning of these words wherever possible. So you will be glad that you can represent the “possessing” of an object as a literal, physical “sitting upon” it (possess = _potis_ + _sedeo_). This is just how the dream-work proceeds. In such circumstances you can hardly demand great accuracy of representation, neither will you quarrel with the dream-work for replacing an element which is difficult to reduce to pictorial form, such as the idea of breaking marriage vows, by some other kind of breaking, e.g.