The falling of light on the closed eyes, or the half opening of the eyes, has been found to serve as a visual stimulus to dreams, but I have myself no decisive evidence on this point.[67] In the case of a lady who dreamed that a lover was in her room, and that suddenly the door opened, and she saw her mother standing before her with a bright light, which awoke her, she could find nothing in the room, no light, to account for the dream. It is, of course, unnecessary for a dream of a bright light to be actually produced by an external visual stimulus accompanying the dream, for the spontaneous retino-cerebral activity itself produces sensations of light. Thus, on the night after a pleasant walk in a country lane through which the setting sun shone, I dreamed that I was walking along a lane in which I saw a bright light and my own vast shadow in front of me. It would seem that, on the whole, the curtain of the eyelids effectually shuts out light from the eye during sleep, and that the sense which is more active during the day than any other is the most carefully guarded of all during the night. The peculiarly delicate and unstable nature of the chemical basis of vision makes up for this protection from external stimulation, and by its spontaneous activity ensures that even in dreams vision is the predominant sense. What we find as regards the part played in dreams by excitations arising from the external specific senses holds good also for excitations arising from internal organic sensations. The main difference is that the stimuli which reach sleeping consciousness from the organs within the body--the stomach, heart, lungs, sexual apparatus, bladder, etc.--are usually more vague and massive, more difficult to recognise and identify, than are the more specific sensory stimuli which reach us from without. These visceral excitations may be transformed within the brain into imagery so unlike themselves that we may refuse to recognise them, and must frequently experience some amount of hesitation. Evidence of this fact will come before us in due course later on. I only wish to refer here to the more obvious part played in dreams by sensations arising within the body. We should expect that the visceral processes to be translated most clearly and directly into dreaming consciousness would be, not those which are regular and continuous, but those which assert themselves, more or less imperiously, at intervals. This is actually the case. The heart, for instance, probably plays a part in dreams only when disturbed in its action, and even then nearly always a very transformed part. On the other hand, when the impulses of the generative system arise in sleep to manifest themselves in erotic dreams, the resulting imagery is usually very clear, and with very definite and recognisable sexual associations. Erotic dreams are, indeed, in both men and women, among the most vivid of all dreams, and the most emotionally potent.[68]