into the discussion of these relatively familiar cases, and I shall turn at once to the more complex types.
Automatic writing is an exceedingly common phenomenon. It took its rise from table-turning. Ordinary tables being found in many cases too heavy for the "spirits" to lift, tiny three-legged tables were made for the purpose and termed "planchettes." Later the device was hit upon of attaching a pencil to one leg and placing a sheet of paper beneath to record the movements of the leg. This is our modern planchette. Two or three persons then put their hands on the instrument and wait to see "what planchette will say." Many automatists need no planchette. It is enough for them to take a pencil in hand and sit quietly with the hand on a sheet of paper. After the lapse of a variable period of time the hand will stiffen, twist, and fall to writing quite of its own accord. Of these methods planchette is the more likely to be successful. In the first place, the chances of finding an automatist among two or three people is obviously greater than in the case of one; furthermore, since all expect planchette to move, the slightest tendency to automatism on the part of any one is likely to be magnified by the unconscious co-operation of the others, and is less likely to be checked by the writer himself, since each ascribes the movement to any one but himself.
The writing produced by either of these methods may be regarded as belonging to one of two main types: 1. That which, although involuntary, is dependent upon the co-operation of the subject's consciousness. 2. That which is produced without the co-operation of the subject's consciousness. The latter, again, may be either intelligible or in "unknown tongues."
Intelligible automatic writing may be produced without the co-operation of the subject's consciousness, either when that consciousness is apparently unimpaired, or when the patient is in a trance state. The latter I need not now discuss, as it belongs to the same category as dreams, but the former calls for some comment.
There are two methods of proving that the automatic messages did not emanate from the subject's upper consciousness. In the first place, it is sometimes found that they become the more clear and copious the more effectually the upper consciousness of the subject is distracted from the writing. Miss G, for example, whom I studied with some care, always did her best automatic writing when busily engaged in conversation or in reading aloud. I concealed her hand from her eyes, and it was but now and then that she would decipher a word by the sense of touch and movement as it was written. But the messages she wrote were always trivial, silly, and often self-contradictory.
In the second place, the content of the writing may be of such
