I have already described how, by concentrating the mind upon the subject, these dreams can be made to recur, and how the accomplishment of definite acts of flight and new methods of flying can be acquired by the exercise of will-power.

On the other hand, I have made very many attempts to see whether some particular bodily attitude during sleep would have any effect upon dream flying. My notes show that lying on my back, lying upon either side, or as I often lie face downwards with my head pillowed on one arm, the sleeping attitude apparently makes no difference whatever to the flying dream. I made other experiments to see whether a pillow wedged at the foot of the bed, so that the feet were firmly pressed against it, would make any difference, but it did not apparently do so. The hands, moreover, are constantly used in these dreams, a slight waving or paddling motion giving direction or velocity to the flight; but I find that I fly just as well if either or both my hands are imprisoned under the body, and even the fact of waking up with one hand "gone to sleep" by reason of continued pressure on it has not apparently affected the dream.

The conclusion that I gather from these experiments is that, whilst by an act of will it is possible to recapture the flying dream, and whilst the memory of past acts of flight helps to make such dreams recur, I can find little confirmation of the theory that these particular dreams are caused by physical sensations. The question of dream control has, however, been fully discussed and only belongs very indirectly to the question now under consideration of the part that sensory impressions play in the causation of our dreams.

Having read of experiments being made to test whether dreams are directly affected by the senses of smell and taste, I tried certain of these experiments, and for this purpose kept on my pillow at night a few lavender flowers or rose-leaves, or a tiny grain of camphor or spice; but whilst the presence of one of these scented things leaves my dreams untouched, it will produce a most vivid impression in the moments between sleeping and waking. This intensification of the senses when sleep is approaching has probably often been observed. The subject belongs more properly to a later chapter. Here I need only say that whilst at these times the senses of smell and taste become abnormally acute, these sense impressions do not"carry on" into the dreams that follow, and do not apparently influence them in any way. Incidents often occur in dreams in which I enjoy something that has a delicious flavour, but these incidents seem always to spring from memories of taste, and can be traced back to such memories, but not to any actual or present sensation; for even when a morsel of spice has been kept in the mouth during sleep it has not succeeded in affecting the dream.

The sense of warmth and cold seems to affect dreaming more than any other direct sense impression, and both the nature of dreams and the power of remembering and recording them are influenced by the degrees of bodily warmth during sleep.

It is a fact for which there is probably some quite simple physical explanation, that in order to be able to recall long dreams easily and accurately the body must be kept fairly and evenly warm; and as I personally like to sleep under only very light coverings, I have learned that I must add to these if the best dreams are to be dreamed or recorded. Delightful dreams are more apt to recur, and are more apt to be vivid in character, if the body is kept fairly warm during sleep. On the other hand, many people bear testimony to the fact that sleeping under an eiderdown quilt or any coverings that produce undue warmth "give bad dreams" to the sleeper. I do not know of any other purely physical means by which the nature of dreams can be thus easily altered, but by variations in bodily temperature I have found that they can be modified, and the recording of long and complicated dreams is made easier to me if that temperature is rather carefully regulated.

From whatever side we approach the subject of dreams there seems to be endless variety in individual experience concerning them, and on no point do dreamers seem to differ more widely than as to the memory that they retain in dreams of sense impressions.

To those who have a very complete and full dream life, and for whom these experiences make up no inconsiderable part of the pleasure of living, all sense impressions will naturally take their due place during sleep. The memory of colour, light, and sound, of fine taste and delicate scent, and all the delight that our keenest sense impressions give us, should naturally be found repeating themselves in dreams; but in this respect, as in so many others, there is a surprising diversity in people's experiences. It is interesting to note the answers that we get if we put such questions to our friends as: "Do you see colour in your dreams? Is your dream world of varied colour like our own, or is it neutral tinted?" Although, I suppose, we all spend a considerable proportion of our life in dreaming, it is curious to find how many people, when they are asked such a question, cannot answer it, or recall with clearness anything about these aspects of their dream life. Some people, however, have told me at once that either they see no colour at all in dreams or that colour has made so little impression on them that they have no recollection of it.

A few have said that their dream world is definitely colourless like a monochrome drawing, and that there seems to be very little bright sunlight or deep shadow in it. Others, again, say, and say most convincingly, that they see everything with the colours, lights, and shadows that we see in the world of day; but perhaps the commonest experience is to be able to recall very clearly some one passage or note of colour in a dream, whilst all recollection of colour in the rest of it has passed away. In nearly all my own dreams I find this to be the case; the one colour note that is recalled is generally vividly retained, but the other objects of the dream are not remembered in colour at all, colour not being the fact about them that has arrested the attention enough to be remembered.

When in after years I read in Robert Louis Stevenson's "Chapter on Dreams" that as a child he had been "haunted by nothing more definite than a certain hue of brown," which he did not mind in the least when he was awake, but feared and loathed while he was dreaming, it recalled vividly a dream which used from time to time to haunt my own child-hood. At a spot in Kensington Gardens which is now dominated by Watts' statue of "Physical Energy," the dream used to begin. The slow ticking of a watch gave warning, but alas! too late, of what was inevitably approaching. In the eastern sky a cloud would come slowly sailing up till it reached and covered the zenith, a cloud of a shade of purplish brown which for some unknown reason was dreadful to me, and lastly, from somewhere very far away there would come the sound of the roll of drums. I thought life could hold nothing more full of fear than that colour accompanied by that distant sound, and no daylight reasoning gave me reassurance, until by a happy chance I learned how to make my escape from all such haunting dreams.

A statement that appeared some years ago in a medical journal to the effect that colour vision in dreams is more often than not either imperfect or altogether absent, suggested this point to me as one that might well be noticed and recorded. I showed my notes to my father and asked him about his own memory of colour.

He said that colour was always a prominent thing in his dreams. He instanced one favourite dream that from an early age had recurred from time to time during his long life. In it he would find himself on the top of the stage-coach, on which as a little boy he used to travel part of the lengthy journey between his school and his home, but the dream journey took him through country unknown to him, and of singular beauty. He described the gradual descent of the coach into a wide and fair valley; the colour of the trees on either side and the exquisite blue of the far distance that could be seen beyond the furthest limits of the valley were the elements in the dream scene that had made the greatest impression on his memory. The subject having been thus suggested, he told me of other dreams in which colour vision was a prominent feature. Colour sense was evidently strong in them, and often seemed to be the point that most impressed his memory. The blossoming and vivid colour of his favourite crimson hawthorn-tree in a dream was reported to me one May morning during the last week of his life.

My mother's dreams also seem to centre round colour, and colour seems to be the fact in them that she remembers most clearly.

"In a gentian dream last night," she records, "I found a lovely clump of Gentiana verna growing on low rocks together with a mossy-looking plant whose yellower green was in strong contrast with the colour of the leaves of the gentian - the starry flowers were, of course, of the deep gentian blue".

Just as some people describe their dream world as being a colourless one, others say that they have no sense of smell in their dreams.1 Possibly it is their memory of these sense impressions that is at fault, for I find in my own notes many observations of smell as well as of taste, observations not only of fragrant flower scents, but also of subtler impressions of smell such as that which occurs in the following dream: "I was motoring through a countryside of steep hills and valleys; dusk was beginning to; fall; 'lighting-up time' had not actually come, but I was warned by the faint scent, cold, clean and unmistakable, that belongs to valley mist, that, besides the natural darkening of the evening, a light mist was beginning to rise and to obscure the road".

The sense of taste comes also into certain dreams, but all these impressions are quickly forgotten when we wake. I have looked at the wares of many a confectioner, in the hope, never, I fear, to be realised in a waking world, that I might find amongst them the candy of a certain dream, candy of an entrancing green colour, and of a flavour that only a dream confectioner seems able to supply.

1 "Very infrequently do we dream words, and almost never do we dream smells or tastes."- Lay, "Man's Unconscious Conflict," p. 167.