[27] This is the explanation offered by, for example, Delage (Comptes-rendus de l'Académie des Sciences, vol. cxxxvi., No. 12, pp. 731 et seq.). It is accepted by Guyon and others. Delage insists on the retinal element since he finds that hypnagogic images follow the movements of the eye.
[28] Similarly, under chloroform, Elmer Jones found that vision is at first stimulated.
[29] G. H. Savage, 'Dreams: Normal and Morbid,' St. Thomas's Hospital Gazette, February 1908.
[30] British Medical Journal, 11th May 1907. The actual hallucinations of the insane are usually coloured normally. Head, however, finds (Brain, 1901, p. 353) that the waking visual hallucinations sometimes associated with visceral disease are always white, black, or grey, and never coloured or even tinted.
[31] The transformation of birds into human beings seems peculiarly common in dreams. I have referred to this point elsewhere (Studies in the Psychology of Sex, vol. i. 3rd ed., p. 193). It is an interesting and doubtless significant fact that the same transformation is accepted in the myths of primitive peoples. Thus, according to H. H. Bancroft (Native Races of the Pacific, vol. i. p. 93), a pantomime dance of the Aleuts represents the transformation of a captive bird into a lovely woman who falls exhausted into the arms of the hunter.
[32] It is noteworthy that this marked tendency in dreams to discover analogies, although doubtless a tendency of primitive thought, is also a progressive tendency. 'The conquests of science,' says Sageret ('L'Analogie Scientifique,' Revue Philosophique, January 1909), 'are the conquests of analogy.'
[33] Maury, Le Sommeil et les Rêves, p. 115.
[34] Kraepelin, 'Ueber Sprachstörungen im Träume,' Psychologische Arbeiten, Bd. v., 1906, pp. 1-104; cf. Lombard, 'Glossolalie,' Archives de Psychologie, July 1907.
[35] This is confirmed by the fact that under chloroform anaesthesia hearing is the first sense to be lost and vision the last (Elmer Jones, 'The Waning of Consciousness under Chloroform,' Psychological Review, January 1909).
[36] It may be recalled as not without significance that the formation of new words is fairly common among young children; see, e.g., an interesting correspondence in Nature, 26th March and 9th April 1891.