psychic process. The old simile of "the ten fingers of a person ignorant of music running over the keyboard of an instrument" perhaps best illustrates in what esteem the dream is commonly held by the representatives of exact science. Thus conceived, it becomes something wholly insusceptible of interpretation. How could the ten fingers of a player ignorant of music perform a musical composition?

The theory of partial wakefulness did not escape criticism even by the earlier writers. Thus Burdach wrote in 1830: "If we say that dreaming is a partial waking, then, in the first place, neither the waking nor the sleeping state is explained thereby; secondly, this amounts only to saying that certain powers of the mind are active in dreams while others are at rest. But such irregularities occur throughout life..." (p. 482).

The prevailing dream-theory which conceives the dream as a "physical" process finds a certain support in a very interesting conception of the dream which was first propounded by Robert in 1866, and which is seductive because it assigns to the dream a function or a useful result. As the basis of his theory Robert takes two objectively observable facts which we have already discussed in our consideration of dream-material (chapter I., B). These facts are: (1) that one very often dreams about the most insignificant impressions of the day; and (2) that one rarely carries over into the dream the absorbing interests of the day. Robert asserts as an indisputable fact that those matters which have been fully settled and solved never evoke dreams, but only such as lie incompleted in the mind, or touch it merely in passing (p. 10). "For this reason we cannot usually explain our dreams, since their causes are to be found in sensory impressions of the preceding day which have not attained sufficient recognition on the part of the dreamer." The condition permitting an impression to reach the dream is, therefore, that this impression has been disturbed in its elaboration, or that it was too insignificant to lay claim to such elaboration.

Robert therefore conceives the dream "as a physical process of elimination which in its psychic reaction reaches the consciousness." Dreams are eliminations of thoughts nipped in the bud. "A man deprived of the capacity for dreaming would in time become mentally unbalanced, because an immense number of unfinished and unsolved thoughts and superficial impressions would accumulate in his brain, under the pressure of which all that should be incorporated in the memory as a completed whole would be stifled." The dream acts as a safety-valve for the over-burdened brain. Dreams possess a healing and unburdening power (p. 32).

We should misunderstand Robert if we were to ask him how representation in the dream could bring about an unburdening of the mind. The writer apparently concluded from these two peculiarities of the dream-material that during sleep such an elimination of worthless impressions is effected somehow as a somatic process; and that dreaming is not a special psychic process, but only the information which we receive of such elimination. Moreover, elimination is not the only thing that takes place in the mind during sleep. Robert himself adds that the stimuli of the day are likewise elaborated, and "what cannot be eliminated from the undigested thought-material lying in the mind is bound up into a completed whole by mental clues borrowed from the imagination, and is thus enrolled in the memory as a harmless phantasy- picture" (p. 23).

But it is in his criticism of the sources of dreams that Robert is most flatly opposed to the prevailing theory. Whereas according to this theory there would be no dream if the external and internal sensory stimuli did not repeatedly wake the mind, according to Robert the impulse to dream lies in the mind itself. It lies in the overloading of the mind, which demands discharge, and Robert considers, quite consistently, that those causes conditioning the dream which depend on the physical condition assume a subordinate rank, and could not incite dreams in a mind which contained no material for dream-formation derived from the waking consciousness. It is admitted, however, that the phantasy-images originating in the depths of the mind may be influenced by nervous stimuli (p. 48). Thus, according to Robert, dreams are not, after all, wholly dependent on the somatic element. Dreaming is, of course, not a psychic process, and it has no place among the psychic processes of the waking state; it is a nocturnal somatic process in the apparatus of mental activity, and has a function to perform, viz., to guard this apparatus against excessive strain, or, if we may be allowed to change the comparison, to cleanse the mind.