symptoms in the psychoneuroses. In brief, the relation which we are here investigating is by no means an exclusive one; the fact remains, however, that this characteristic of the dream, whenever it occurs, seems to be its most noteworthy characteristic, so that we cannot think of the dream-life without it. To understand it, however, requires a very exhaustive discussion.

Among all the observations relating to the theory of dreams to be found in the literature of the subject, I should like to lay stress upon one as being particularly worthy of mention. The famous G. T. H. Fechner makes the conjecture, * in a discussion as to the nature of the dreams, that the dream is staged elsewhere than in the waking ideation. No other assumption enables us to comprehend the special peculiarities of the dream- life.

* Psychophysik, Part. II, p. 520.

The idea which is thus put before us is one of psychic locality. We shall wholly ignore the fact that the psychic apparatus concerned is known to us also as an anatomical preparation, and we shall carefully avoid the temptation to determine the psychic locality in any anatomical sense. We shall remain on psychological ground, and we shall do no more than accept the invitation to think of the instrument which serves the psychic activities much as we think of a compound microscope, a photographic camera, or other apparatus. The psychic locality, then, corresponds to a place within such an apparatus in which one of the preliminary phases of the image comes into existence. As is well known, there are in the microscope and the telescope such ideal localities or planes, in which no tangible portion of the apparatus is located. I think it superfluous to apologize for the imperfections of this and all similar figures. These comparisons are designed only to assist us in our attempt to make intelligible the complication of the psychic performance by dissecting it and referring the individual performances to the individual components of the apparatus. So far as I am aware, no attempt has yet been made to divine the construction of the psychic instrument by means of such dissection. I see no harm in such an attempt; I think that we should give free rein to our conjectures, provided we keep our heads and do not mistake the scaffolding for the building. Since for the first approach to any unknown subject we need the help only of auxiliary ideas, we shall prefer the crudest and most tangible hypothesis to all others.

Accordingly, we conceive the psychic apparatus as a compound instrument, the component parts of which we shall call instances, or, for the sake of clearness, systems. We shall then anticipate that these systems may perhaps maintain a constant spatial orientation to one another, very much as do the different and successive systems of lenses of a telescope. Strictly speaking, there is no need to assume an actual spatial arrangement of the psychic system. It will be enough for our purpose if a definite sequence is established, so that in certain psychic events the system will be traversed by the excitation in a definite temporal order. This order may be different in the case of other processes; such a possibility is left open. For the sake of brevity, we shall henceforth speak of the component parts of the apparatus as Psi-systems.

The first thing that strikes us is the fact that the apparatus composed of Psi-systems has a direction. All our psychic activities proceed from (inner or outer) stimuli and terminate in innervations. We thus ascribe to the apparatus a sensory and a motor end; at the sensory end we find a system which receives the perceptions, ind at the motor end another which opens the sluices of motility. The psychic process generally runs from the perceptive end to the motor end. The most general scheme of the psychic apparatus has therefore the following appearance as shown in Fig. 1. (See illustration.) But this is only in compliance with the requirement, long familiar to us, that the psychic apparatus must be constructed like a reflex apparatus. The reflex act remains the type of every psychic activity as well.

We now have reason to admit a first differentiation at the sensory end. The percepts that come to us leave in our psychic apparatus a trace, which we may call a memory-trace. The function related to this memory-trace we call the memory. If we hold seriously to our resolution to connect the psychic processes into systems, the memory-trace can consist only of lasting changes in the elements of the systems. But, as has already been shown elsewhere, obvious difficulties arise when one and the same system is faithfully to preserve changes in its elements and still to remain fresh and receptive in respect of new