688 Proceedings. 



concealed problem. Of these, the last two were the more complex, 

 and in the fourth case the ordinary line of operation in waking-hours 

 was entirely reversed. In the second class, the introduction of a second 

 personality was so similar to the common illusions of sleep that classes 

 1 and 2 might fairly be taken as different forms of the same phenomenon. 

 It was not so easy, however, to account for the illusion of debate, where 

 the adversary often replies with a quite unexpected argument, which the 

 sleeper finds it difficult to meet. As an example of the illusion of reading, 

 he only lately imagined that he was reading a certain newspaper (of 

 which, by the way, he had never seen but one copy). The dream paper 

 corresponded perfectly with the real one ; but contained two printers' 

 errors so odd that he (as was his custom in such cases) made a note of 

 them. He remembered both on waking. One was in an article on 

 homceopathy, where the compositor, mistaking the word " dilution " for 

 a collocation of numerals, had set it up " DI LV XIII." Such an error, 

 though possible, was in the last degree improbable. The other he had for- 

 gotten. By what strange process were these blunders at once invented 

 and presented to the dreamer with a full sense of their grotesque incon- 

 gruity ? As for the solving of problems, he had repeatedly had "nuts 

 to crack" in sleep, had succeeded, and had felt real pleasure when the 

 answer flashed upon him. Did he at the same time unconsciously con- 

 struct the problem? The most sustained effort of the kind that he could 

 remember was a double acrostic in verse which he imagined he was read- 

 ing in a magazine. This he studied, word by word, until he succeeded in 

 working out the whole. He was unable to recall all the details on waking, 

 and problem and solution may have been alike defective. Not having 

 noted it at the time, lie could only now remember that one of the key- 

 words was " Oriole " — exactly such a word as is chosen by the construc- 

 tor of this kind of riddle. Had he set himself that puzzle ? He was 

 aware that much had been written upon the main question ; and he 

 had no theory to propound. What he maintained was that the 

 phenomenon of dreaming did not stand by itself, and could not be 

 accounted for in the easy off-hand way in which many writers on 

 popular science dealt with it. An inquiry of this kind was not a mere 

 matter of idle curiosity ; for, as the mental machinery was for the time 

 outside of voluntary control, the phenomena of dreams might enable 

 some idea to be formed of the operations of the mind in certain forms 

 of insanity, whether induced by ordinary disease or by specific nerve- 

 poisons such as alcohol, cannabis, or morphia. The paper thus concluded : 

 " Any theory of mental operations, conscious or unconscious, must be, 

 to borrow a term from the botanist, either endogenous or exogenous. 

 I incline to the latter. There are phenomena seemingly suggestive 

 of a transference of thought or perception which may be illustrated by 

 what is known as induction in telegraphy, when the current jumps from 

 wire to wire, and the message is recorded by the wrong instrument. Such 

 a suggestion fails to meet the cases I have indicated. It is, I think, the 

 more reasonable view that our most original thoughts and inventions, 

 sleeping or waking, orderly or otherwise, reach us from an exterior sphere. 

 Where the soil has, by thought and training, been duly prepared, the 

 living germs fall, and in due time bring forth their fruit, good, bad, or 

 indifferent, each after its kind. Who has not felt a thought flash on him 

 from without, impressing itself like a lightning photograph ? Has not 

 the afflatus of the poet always been recognised as an inspiration ? And 

 even the broken and imperfect phenomena of dreams may throw light on 

 the obscurer operations of the human mind." 



